Data Center Tier Levels are a standard methodology used to determine a data center’s ability to maintain functionality during different types of failures, such as power outages. The higher the tier level, the more sustainability for data center operations and fault-tolerance systems that will allow for efficient use during certain kind of emergencies.
Developed by the Uptime Institute - a global research organization - the tiered system offers organizations a way to evaluate return on investment (ROI) and performance. It is comprised of a 4-tiered scale, with Tier-IV being the most robust.
Uptime Institute is responsible for providing certificates to data center facilities.
The 4 tiers as categorized by the Uptime Institute are:
Tier-I: It’s composed of a single, non-redundant path for distribution, power and cooling with providing 99.671% uptime.
Tier-II: Composed of a single, redundant path for distribution, power and cooling with providing 99.741% uptime.
Tier-III: It has typically more comprehensive protection for power outages and have what’s called N+1 redundancy, a reliable backup power system.
Tier-IV: It tops the level. It represents a data center facility that has the infrastructure, capacity, and processes in place for providing maximum level of uptime. Apart from all requirements for Tiers I, II and III, Tier-IV requires the infrastructure that is fully fault-tolerant. It should function normal even in the event of failure of one or more equipment. It will have multiple cooling units, backup generators, power sources, chillers etc. In the event of failure of one of the equipment, another will start up replacing its output instantly.
What are data center tiers?
Tiers are levels for classifying data centres into 4 categories indicative of their reliability. This four-tier rating system begins with tier-1 data centers that offer no redundancy to tier-4 data centers that offer 2N+1 redundant infrastructure and are completely fault tolerant with no single point of failure.
Tier-1 data centers cannot have more than 28.8 hours of downtime annually.
Tier-2 data centers are partially redundant in terms of power and cooling and cannot have more than 22 hours of downtime every year.
Tier-3 data centers are usually sought by medium-and-large-sized businesses and enjoy N+1 redundancy. They cannot have more than 1.6 hours of downtime annually.
Tier-4 data centers typically cater to enterprises and can have a maximum of 26.3 minutes of downtime every year. They have a completely redundant infrastructure.