Most modern businesses would cease to exist without data centres. The computing power they provide is the backbone of many businesses’ digital infrastructure, from backing up files to web hosting to powering digital services. All of this highlights just how important data centres are to modern life – and how important it is that you pick the right one.

When choosing a data centre, there is no more reliable or popular metric than The Uptime Institute’s Tier Classification System. In place for almost 30 years, the Tier Classification System remains the international standard for data centre performance. Here’s a rundown of what the system is, how it works, and why it’s such an important tool for classifying data centres.

What is the Tier Classification System?

The Uptime Institute created the Tier Classification System in the 1990s to provide a common language for evaluating data centre performance. Over the years, it has evolved to become a global industry standard for both the infrastructure and operation of data centres. The Tier Standard shows data centres the standards they need to meet to improve their design topology, in order to improve both infrastructure performance and server uptime.

For anyone who’s looking for a data centre, the Tier Classification System is an easy way to determine the reliability of a data centre’s infrastructure. The ranking primarily reflects the likely uptime of the data centre, but also its ability to withstand disruption, staff expertise, power infrastructure, and the cost of the service based on these factors. Ratings are optional, but most data centres seek a rating from The Uptime Institute, as a good rating reflects well on the business.

How do data centre tiers work?

The Uptime Institute classifies data centres by placing them in one of four Tiers:

• Tier I – Basic Capacity
• Tier II – Redundant Capacity
• Tier III – Concurrently Maintainable
• Tier IV – Fault Tolerant

Each tier defines the criteria data centres need to meet in areas such as maintenance, power, cooling and fault resilience. Each successive Tier includes the requirements of the previous Tier, along with a host of new requirements. What’s important to note is that a Tier IV data centre is not immediately or unquestionably better than a Tier I or II. Instead, they meet different requirements for different business situations and use cases.

Each Tier consists of two key components: topology and operational sustainability. Topology reflects the various aspects of data centre infrastructure that are required to ensure resilience, capacity, availability and performance. Operational sustainability relates to the business practices required to implement these topological factors successfully, and ensure that the requirements of a Tier are both achieved and maintained.

What does each data centre tier mean?

The data centre tiers can be broken down as follows:

Tier I – Basic Capacity

A Tier I data centre offers a basic level of capacity and uptime, suitable for an average office environment. It should protect against brief outages as a result of human error, but is not expected to protect against unexpected failures, such as those caused by long term power outages or equipment failures. A Tier I facility may have to be shut down entirely for maintenance, either in an emergency or preventative capacity.

Requirements for a Tier I data centre include:

• 99.671% guaranteed uptime
• Uninterruptible power supply (UPS)
• Dedicated area for IT systems and equipment
• 24/7 cooling infrastructure
• A working generator for power outages

Tier II – Redundant Capacity

As well as including the requirements from the previous Tier, a Tier II data centre adds additional requirements for redundancy. This primarily provides increased protection against power outages, but also allows for component failures and planned maintenance. The idea is that aspects of the data centre can be shut down, removed and replaced without having to shut down the entire data centre.

Requirements for a Tier II data centre include:

• 99.741% guaranteed uptime
• Redundant power generation (e.g. generators)
• Redundant power storage (e.g. batteries, fuel cells)
• Redundant cooling systems (e.g. pumps, chillers)

Tier III – Concurrently Maintainable

As well as including the requirements from the previous Tier, a Tier III data centre is distinguished by having redundant components that can be swapped in and out without affecting capacity or uptime. In essence, a Tier III data centre should be able to undertake planned maintenance or repairs without having to shut anything down, by virtue of having sufficient redundancy in its equipment.

Requirements for a Tier III data centre include:

• 99.982% guaranteed uptime
• Full component redundancy
• Partial concurrent maintenance

Tier IV – Fault Tolerant

As well as including the requirements from the previous Tier, a Tier IV data centre adds complete redundancy of equipment. In a Tier IV data centre, both capacity and distribution equipment should be able to fail without causing downtime in all but the most extreme circumstances. This provides not just redundancy in cases of planned maintenance, but in case of unplanned failures, providing the highest reasonable level of service for clients.

Requirements for a Tier IV data centre include:

• 99.995% guaranteed uptime
• Fault tolerance
• Full concurrent maintenance
• 24/7/365 staffing

–

The Tier Classification System is a helpful guide to data centre performance and resilience, and a reliable tool for organisations for more than 30 years. What’s important to remember is that Tier IV isn’t always better than Tier I – instead, the data centre tier you choose will be dictated by your requirements.

Any data centre with a Tier Classification has met exacting standards, and will be a good choice for applicable businesses. Sota utilises multiple Tier Classified data centres and global transit providers as part of our fully integrated network. To learn more about our resilient and reliable services, get in touch with us today.