In 1995, The Uptime Institute (TUI) developed an objective, empirically based, four tier classification system to assess a data center’s quality and reliability. The purpose of TUI’s four tier system is to provide companies searching for datacenter space with simple, non-technical means for evaluating the differences in datacenter investments. Prior to the advent of TUI’s four tier classification system, there were no simple, objective standards for evaluating a data center’s functionality and sustainability. Thus, TUI provides the IT industry a vital, third-party assessment of a data center’s site infrastructure topology. TUI’s classification system is broken down into four rigorously defined tier levels. Even though a data center will need to fulfill the basic minimum requirements of a given tier, each tier level builds off of the previous tier’s design topology. For example, a tier three data center will not only meet the minimum requirements of being a tier three data center, but it will also have all of the components of a tier two and tier one data center. Here are the four tiers:
TIER 1 Data Center
BASIC REQUIREMENTS:
- Non-redundant capacity components
- Single non-redundant distribution paths
The first and most basic type of data center is a tier one data center. This type of data center has the most basic site infrastructure. A tier one data center has non-redundant capacity components, and single non-redundant distribution paths. In addition, a tier one datacenter is not concurrently maintainable, and is not fault tolerant. Due to not being concurrently maintainable, a tier one datacenter must be completely shut off during all planned and unplanned maintenance and repair work. Furthermore, any capacity component or distribution path failure will impact the computer systems and may result in the shutdown of the site infrastructure, which will invariably affect the end user.
TIER 2 Data Center
BASIC REQUIREMENTS:
- Redundant capacity components
- Single non-redundant distribution paths
The primary advantage and fundamental requirement for a data center to be considered a tier two data center are its redundant capacity components. This ensures that a capacity component failure MAY only disrupt the computer equipment; unlike a tier one data center, where a capacity component failure will always impact the computer equipment. Note, however, similar to a tier one data center, a tier two data center is still susceptible to distribution path failures, site infrastructure shutdowns from planned and unplanned maintenance, operational errors, and spontaneous failures of site infrastructure components.
TIER 3 Data Center
BASIC REQUIREMENTS:
- Redundant capacity components
- Multiple distribution paths (Alternate)
- Concurrently maintainable
The fundamental requirement for a tier three data center is that it needs to be concurrently maintainable. “Concurrent maintainability” means that every capacity component and constituent of the power distribution paths can be removed without any of the computer equipment shutting down. A tier three data center is concurrently maintainable by implementing redundant capacity components, employing multiple distribution paths serving the site infrastructure’s computer equipment, and applying dual power inputs between the UPS (uninterruptable power supply) and the computer equipment. It is important to note that although a tier three data center has multiple distribution paths, usually only one of the distribution paths is active and serving the site infrastructure’s computer equipment at any one time. Similar to tier one and tier two data centers, tier three data centers are also susceptible to operation errors and spontaneous failures of site infrastructure components.
TIER 4 Data Center
BASIC REQUIREMENTS:
- Redundant capacity systems
- Multiple distribution paths (simultaneous)
- Dual powered IT equipment
- Fault tolerant
A tier four data center is the most robust of all data centers. It is fault tolerant, which means it can sustain a single, worst-case, unplanned event while not disrupting the end user. In addition, a facility failure will not impact the computer equipment at the site infrastructure. A tier four data center is fault tolerant by physically separating its redundant capacity systems and distribution paths. Unlike the alternate distribution paths that are implemented in tier three data centers, a tier four data center implements simultaneous distribution paths. A tier four simultaneous distribution path will provide more protection against an unplanned distribution path failure than will the tier three alternate distribution paths. Simultaneous distribution paths provide multiple power avenues to the computer equipment at any given time. This ensures that if there is a single, worst-case distribution path failure, the site’s computer equipment will still be up and running without any downtime. A tier four data center is not susceptible to disruption from planned maintenance or repair activities, although, the risk of disruption may still be elevated during these times. A tier four data center may still be disrupted by the EPO (emergency power off), or multiple failures at the facility.
Provided by Facility Gateway Corporation.
Article Sources: