About Data Centers

As Internet website hosting and application hosting continues to grow as a major industry, many competing companies offering data centers with "100% uptime" all actually have different infrastructure capabilities. How are different data center infrastructures distinguished? What are the major data center types? What kind of data center is right for your business?

Within the data center space, the ultimate goal is to achieve "Five Nines". In other words, "Five Nines" corresponds to 99.999% uptime. Distinct data center infrastructure designs can accomodate for different levels of uptime based on business needs and requirements. Regardless, the highest level data center (Tier IV, which is fault tolerant and concurrently maintainable) can not achieve this and only delivers a 99.995% average uptime.

Data center designs have evolved through four distinct stages, which are classified into separate tiers. Each tier constitutes an increase in reliability, power paths, redundancy and availability. The following list summarizes the characteristics of each data center tier.

Tier I Data Center - Single path for power and cooling distribution with no reduntant components - 99.671% availability. A Tier I data center is susceptible to disruptions from both planned and unplanned activities. Data center disruptions are caused by operation errors or spontaneous failures of site infrastructure components. Tier I data centers should be shut down at least once per year for repairs and maintenance.

Tier II Data Center - Single path for power and cooling distribution with redundant components - 99.741% availability. Because of its redundant components, there is less risk of disruptions from scheduled and unpredictable activity compared to Tier I data centers. Tier II data centers are designed with a single-threaded distribution path throughout. This translates into a "Need plus One" capacity (N+1). If and when critical power paths and other site infrastructure require maintenance, Tier II data centers must also be shut down.

Tier III Data Center - Multiple power and cooling distribution paths. These data centers only have one active path with redundant components. They are also concurrently maintainable - 99.982% availability. Planned site maintenance can be performed on Tier III data centers without causing any disruptions. Thus, preventative maintenance, testing and repair can be scheduled and carried out without shutting down the infrastructure. Similiar to Tier II data centers, Tier III data centers do have N+1 capacity design. However, due to their multiple paths for power and cooling, annual IT downtime is significantly lower. Unplanned activities like errors in operation and chance failures of infrastructure components can and will result in loss of uptime. When additional protection is warranted, Tier III sites are typically upgraded to Tier IV data centers.

Tier IV Data Center - Multiple active power and cooling distribution paths, redundant components and fault tolerance - 99.995% availability. Unlike a Tier III data center, both distribution paths are active in Tier IV data centers. Because of this fault tolerant functionality, a Tier IV data center provides the capacity to conduct any planned activity without disruption. One unplanned failure can be sustained with no impact as well. Active distribution paths are in place to support this protection in a System+System configuration (S+S). FOr businesses with IT requirements for reliability, availability and serviceability, Tier IV site infrastructures are ideal. Such IT concepts typically entail CPU clustering, RAID DASD adn redundant communications.

Are you interested in finding out what kind of data center fits your business requirements? Please feel free to provide Sterling Network Services with some of your company and contact information and request a quote. You can also contact us directly at (800) 930-1802 to answer some of your more specific questions.