Yet tier certification ratings alone only tell you part of the story. True, they let you compare expected uptime between data center facilities. However, availability is only one piece of the puzzle. You also need to know how to align ratings with different types of business.
In this article, we describe tier classification levels. We position them for different needs. As well as telling you what they can do, we also tell you what they cannot do. You get the whole story about the data center tier system and how to choose for your organization.
All about availability
Businesses rely more and more on their data and information systems. Data center availability is a key factor. The time you can survive without your data or systems depends on your type of business. For some, an outage of days or hours might be tenable. For others, even minutes of downtime could be a disaster. Many businesses therefore look to data center operators for uptime guarantees.
But what sort of uptime promise can you get? How can you compare one specification with another? Enter data center tier classification. Using open or industry standard criteria, an independent entity can assess and rate data centers. A data center can then present its rating to users to set their expectations.
An example of such an entity is the Uptime Institute. It has defined its own data center classifications that are divided into four levels or tiers. The tiers define criteria for maintenance, power, cooling, and fault handling. Each tier builds on the tier below, adding stricter criteria. An example of an open standard is ANSI/TIA-924-B. There are similarities between this standard and the Uptime Institute criteria.
Getting certification for a data center can be expensive. This is especially true for higher tiers like Uptime Institute Tier III and Tier IV. Some data centers may simply declare that their facilities have uptime to the same level as a certain tier. For example, “Tier III equivalent”. However, if no official rating exists, users should check for themselves.
Data Center Tiers I to V
The main differences between data center tiers are in uptime, redundancy, and paths for power. For data center operators, added factors are costs and times to implement. Each tier rating is based on the weakest component or system in the data center. The higher the availability required, the higher the cost of the data center facility.
The Uptime Institute tiers for data centers can be summarized as follows.
A Tier I data center has an area for IT systems with dedicated cooling. It has an uninterruptable power supply (UPS). It also has an engine generator for power outages. Tier I protects against human error. It does not protect against unexpected failure or downtime. The facility must shut down for maintenance and repairs.
A Tier II data center includes Tier I capabilities. It adds redundant power and cooling at the component level. This improves protection against disruptions. Maintenance of components can be done without an entire shutdown. However, failures will affect the system.
A Tier III data center builds on Tier II capabilities. It adds distribution path redundancy to the Tier II component redundancy. No shutdowns are needed for maintenance or replacement of equipment. Any part can be shut down without impact on IT operations.
A Tier IV data center adds fault tolerance to the Tier III capabilities. Redundancy comes from independent and physically isolated systems. There will be no disruption from planned or unplanned events. However, if maintenance is in progress, a failure may mean a higher risk of disruption.
There is also a Tier V. It has been defined by the service provider Switch. Tier V aims to enhance uptime and reliability for colocation facilities. Tier V data centers must meet all the Tier IV conditions plus other even stricter ones.
Which tier rating is right for my business?
How do tier certification ratings relate to the real world? There are no hard and fast rules. However, the following indications may help.