In the Data Center industry, there are certain terms to refer to specific concepts. This is the case of the ‘Tier’, a term used to define the hierarchy which classifies the availability and redundancy of the systems installed in a Data Center. The Tier classification was developed by the Uptime Institute and is based in a 4-scale level: Tier I, Tier II, Tier III and Tier IV.
But, what does this scale answer to? Mainly, the classification refers to aspects such as the time that the systems are functional and without failures, the security of the facilities against natural disasters or cyberattacks, as well as the features of the infrastructure, the materials used or the redundancy of the systems, meaning to have the critical equipment backed up to avoid service interruption when failures happen.
Tier I
This is the lowest level of the scale. Tier I data centers have a more simple infrastructure which allows to have a basic level of service availability. This means it has not a duplication of critical components and can experiment times of inactivity when undertaking maintenance routines or equipment updates. This is a problem for clients in need of strict requirements, such as banks or hyperscalers.
The main benefit of having a Tier I Data Center is the initial cost, which is relatively low. Thus, this type of facility can be used to house non-critical services, that can tolerate periods of inactivity.
Tier II
Tier II Data Centers offer a higher service availability and redundancy than the previous. They have redundant systems to cover certain critical aspects, such as power sources and cooling systems, which allows to take on several maintenance activities without interrupting the activity or by planning short-term service interruptions.
This classification can be functional for non-critical services and applications which require higher levels of reliability but can still tolerate small interruptions.
Tier III
This is the most standard type of Data Center as of today, for the level of service availability and redundancy in relation to the cost of its construction. Tier III facilites have both high redundancy and availability. The systems are duplicated and have independent routes of energy distribution and cooling systems, which allows the maintenance activities without interruptions.
Thus, these facilities have higher availability than the previous two, as they have a high resilience towards failures and a high level of reliability. This is the type more appropriate to host critical services and systems, specially for hyperscalers.
Tier IV
This is the highest level of the scale in terms of availability and redundancy. These facilities have all systems duplicated, including energy distribution routes and cooling systems, having the capacity of resisting failures in any of their components without affecting the service continuity.
Tier IV facilites have the highest level of service availability. They are designed to house critical applications and services, which cannot have any type of interruption. However, these infrastructures have an extremely high cost, compared to the improvement in service that they represent compared to a Tier III Data Center.
How does an organization choose a Tier when building a Data Center?
The election of a Tier level will depend on the specific requirements of each organization in terms of the level of availability that their services need. In general, small enterprises require lesser levels of availability, being able to host their systems in lower Tier infrastructure with a lower cost, whereas critical industries such as banking need top-tier infrastructures, to guarantee a non-interrupted service.
Today, the Tier III standard is the most adequate for a colocation company, in relation to the building and operation costs and the revenues generated. We have both Tier III and IV facilites, with the latest technology, in order to guarantee an uninterrupted services, accordingly to the demands of the digitalized society we live in.