Data Center Tiers Explained: The Role of the Uptime Institute

Data pathways
Data pathways

A data center is a purpose-built facility housing computer systems and associated storage and telecommunication components. It should feature redundant or backup modules and own its independent infrastructure. 

This includes power supply, data communications connections, environmental controls (air conditioning and fire suppression), and appropriate security devices.

Modern data centers are large, industrial-size operations that may consume as much electricity as the average small town. They exist to serve their customers by storing and protecting their data. 

We demystify data standards here through a detailed discussion of data center tiers. But first, we need to set the stage by exploring the broader role of colocation and cloud storage centers. 

What are Data Center Tiers? 

The four levels of data center tiers extend from requirements for a basic server room all the way through to a mission-critical installation such as the NASA control center. 

The highest level is Tier IV. In this case, all subsystems must be fully redundant. 

Moreover, the center must be able to operate independently for an indefinite period in the event of a primary power outage.

Data Center Tier Levels Explained

Overview 

  • Tier 1 – Basic site infrastructure with no redundancy
  • Tier 2 – Tier I plus redundant critical capacity components
  • Tier 3 – Tier II plus maintainable during operation (Concurrently Maintainable)
  • Tier 4 – Critical systems immune from any single fault or failure (Fault Tolerant)

 

Tier 1: Basic Capacity

Tier 1 offers the most basic infrastructure with minimal data center redundancy. They have a single power and cooling path and no backup systems, making them affordable but susceptible to downtime. 

Ideal for small businesses or startups, Tier 1 is cost-effective but involves a higher risk of service interruptions during maintenance or unexpected failures. 

It’s a good fit for organizations that can tolerate occasional outages and have lower uptime requirements. But a word of warning, if any individual components fail here, then your system would fail. 

Tier 2: Redundant Capacity

Tier 2 data centers provide enhanced reliability with redundant power and cooling components. This redundancy reduces the risk of downtime due to component failures. 

While they still have a single power path, backup systems ensure better uptime than Tier 1. Individual components can be lost in these systems without the system going down entirely. 

Suitable for growing businesses needing higher availability, Tier 2 balances cost and improved stability, making it a practical choice for those needing more reliability without a significant cost increase.

Tier 3: Concurrently Maintainable

Tier 3 data centers are designed for continuous operation, allowing for maintenance without downtime. They feature multiple independent power and cooling paths, ensuring service isn’t interrupted during maintenance or failures. 

This makes them ideal for businesses requiring high availability. With a good balance of cost and reliability, Tier 3 facilities are a popular choice for enterprises that can’t afford frequent downtime but don’t require the extreme redundancy of Tier 4.

Tier 4: Fault Tolerant

Tier 4 data centers offer the highest level of reliability and redundancy. They are fault tolerance, which literally means they are tolerant of a fault.

These data centers are designed to handle multiple simultaneous failures (planned or unplanned) and ensure 99.99% uptime with fully redundant power, cooling, and networking components. 

Tier 4 facilities are perfect for mission-critical operations like financial services, healthcare, and large-scale e-commerce, where any downtime is unacceptable.

The extensive redundancy and fault-tolerant design come at a higher cost, but the unparalleled reliability justifies the investment.

When is a Tier 4 not a Tier 4?

Data center tiers are designed to give the full picture on reliability, in a way that discussion of capacity components never could. 

The problem is, systems are only ever as strong as their weakest link. A tier 4 system that goes through to a single breaker, that then goes through to the load, is only ever as strong as that single breaker. The system would effectively be statistically bottlenecked by the breaker. It could have 10 feeds upstream, but if it goes to a single breaker its reliability is entirely governed by the breaker. 

In simple terms, if a facility is almost entirely Tier 4 but one small part is Tier 1, then the whole facility is actually Tier 1. That’s why we always suggest that people exercise caution before talking about capacity components in a vacuum. Because doing so makes the mistake of disregarding compartmentalisation, physical isolation, distribution paths, and the vitally important marriage of electrical and mechanical. 

To determine the real reliability of a system, we need to think about design objectives and functional reliability. Data center tiers provide us with valuable information on which we can determine reliability to do just that. 

Cost Analysis of Each Data Center Tier

The costs associated with different data center tiers vary widely:

  • Tier 1: Low setup and operating costs but higher risk of downtime, suitable for small businesses.
  • Tier 2: Moderate costs with added redundancy, reducing downtime risks, ideal for growing businesses.
  • Tier 3: Higher costs due to concurrent maintainability features, offering a good balance of cost and reliability for enterprises.
  • Tier 4: Highest costs due to full redundancy and fault-tolerant design, ensuring maximum uptime for mission-critical applications.

Choosing the right tier involves balancing budget constraints with the need for uptime and reliability. Understanding these costs helps businesses make informed decisions about their data center investments.

Understanding the Uptime Institute 

The Uptime Institute is a global authority on data center standards. They created the tier classification system to measure data center reliability and uptime. 

This system helps businesses evaluate their data centers, ensuring they meet necessary performance and reliability standards. 

The Uptime Institute conducts thorough assessments and certifications, ensuring that facilities adhere to these rigorous standards, which is crucial for maintaining optimal data center operations.

 

The Role of Seattle-Based Uptime Institute Tiering System

Visionary innovator Kenneth G. Brill founded the Uptime Institute in 1993. He saw the need to encourage improvements in the performance, efficiency, and reliability of business-critical infrastructure through innovation, collaboration, and independent certifications. 

Since then, his initial thinking has developed into globally accepted standard data center tiers for rating the theoretical availability of facilities. Indeed, it can be said he was one of the fathers of mission-critical computing environments, and his development of the Uptime Institute tier system has revolutionized data center facilities.

There are two phases whereby data centers determine their availability and, therefore, the return on investment the design promises. The first phase involves tier certification of the design documents. The second phase ensures that the constructed data center meets these design standards.

Uptime Institute Tiering and Operational Sustainability

Clearly, these certifications are theoretical propositions, assuming all things are equal. In practice, good management and the human factor can mitigate the possibility of achieving these. 

The Uptime Institute tier system adds a layer of operational sustainability certification. This is essential to ensure the design promises are met or exceeded.

This certification addresses management and operational factors. These include staffing and maintenance, the building’s characteristics, and site location (transportation corridors, risk of flooding, etc.).

The three levels of operational sustainability are as follows:

  • Bronze – There are significant opportunities for improvement to achieve the full potential of the installed infrastructure
  • Silver – There are opportunities for improvement to achieve the full potential of the installed infrastructure
  • Gold – The full uptime potential of the installed infrastructure has been realized or exceeded

The overall assessment of a Tier III colocation center, for example, maybe Tier III Silver if it is not achieving all the promises of its design. 

According to the Uptime Institute tiering process, data center tiers encourage an exceptional standard of installed infrastructure and opportunities for improvement. 

Data Centers’ Role in the Evolution of Computing

Data centers appeared in the 1940s to house large, mainframe computers, requiring special environments, including security arrangements, cooling, raised floors, and cable trays.

After microcomputers sprouted everywhere in business, users began acquiring servers to share facilities. Networking evolved and placed these servers in special rooms.

These ‘data centers’, for this is what they were in their infancy, boomed during the 1997 to 2000 dot-com bubble. Companies began building very large facilities to provide commercial clients with a wide range of solutions. 

Private data centers peaked early; then, they began to disappear in the face of special requirements for cloud computing.

These special requirements are security, availability, environmental impact, and adherence to standards in an increasingly challenging environment. 

The Telecommunication Industry Association, accredited by the American National Standards Institute (ANSI), was formed in 1988 and began to establish standards to regulate these matters. 

Nowadays, a modern data center has come a long way from a main-frame room designed on the fly in the 1980s.

Basic Requirements for Modern Data Centers

Colocation and cloud storage centers must enable businesses to meet their stated goals of data reliability and safety. 

They must provide a seamless service, allowing clients to continue operating, even during extreme weather events. Therefore, they must ensure high standards of integrity and functionality.

This involves the redundancy of mechanical cooling and power systems serving the data center along with fiber optic cables. However, this assumes rigorous design and management standards. This is not as common as the industry would expect its customers to expect.

Not all data centers are as good, especially when you can rely on them to be there when you need them. The harsh effects of climate change and extreme weather events add an overburden of concern we can no longer deny. 

We should also no longer accept qualitative promises of ‘very robust’ and ‘extremely unlikely to fail.’ Instead, we should insist on quantitative standards to compare value across the industry.

Telecommunication Industry Association Standards

The Telecommunication Industry Association (TIA) recognized the need for an industry-standard in 2005, the year it published its ANSI/TIA-942, Telecommunications Infrastructure Standard for Data Centers. 

This defined four levels of data centers in a thorough, quantifiable manner. The most recent update appeared in 2017.

Other related TIA standards include:

  • TIA-568-C (Telecommunications Cabling Standards, used by nearly all voice, video and data networks).
  • TIA-569-B (Commercial Building Standards for Telecommunications Pathways and Spaces)
  • TIA-607-B (Commercial Grounding – Earthing – Standards)
  • TIA-598-C (Fiber Optic Color-Coding)  
  • TIA-222-G (Structural Standard for Antenna Supporting Structures and Antennas)
  • TIA-602-A (Data Transmission Systems and Equipment, standardizing the common basic Hayes command set)
  • TIA-102 – (Land Mobile Communications for Public Safety (APCO/P25)

The Telecommunication Industry Association maintains these standards on a participative basis involving engineers, manufacturers, and users. It invites them to submit suggestions and vote on its amendments.

Get More Advice on Data Center Tiers 

Understanding Data Center Tiers is crucial for optimizing your infrastructure and ensuring reliability. 

By familiarizing yourself with these tiers, you can make informed decisions that minimize risks and enhance performance.

For more personalized advice and in-depth guidance, contact us today for more advice today on data center tiers.