Tier Certification system
Are you kidding? This can't be the only commercially available Tier III data center. What about all the data centers
in Atlanta and Charlotte?
No. We're not kidding.
Even though people tend to use the Tier classifications somewhat casually, there are few certified few data center
operations. For one thing, the Certification process alone is expensive. Even more
expensive however are the layers of redundancy for every single component in the center's operations. That means
everything, from the school-bus sized diesel generators to the cloned smoke detectors in the ductwork. Just how much
"everything" encompasses means something very different to you at the end of the process.
"The OnePartner Advanced Technology & Application Center (ATAC) in Duffield, Virginia, is currently the
sole US company providing outsourced
commercial data center services (including colocation and disaster recovery) that has been awarded a Tier III Design Certification by the Uptime
Institute."
-- Julian Kudritzki, Certification manager of the Uptime Institute and
ComputerSite Engineering
Also, there's no such thing as a Tier 2.5 classification. A data center is judged at the level of its weakest
points. If it has dual generators, dual air conditioning units, etc but only one switch for power, it will not
certify higher than a Tier II.
The Certifications are managed by The Uptime Institute. They authorize only one
company (ComputerSite Engineering) to perform the thorough engineering reviews. Any data
center claims for a classification without the actual Certification are immediately suspect.
For reference, the Tier Certification levels can be described as follows:
Tier I
Tier I is composed of a single path for power and cooling distribution, without redundant components, providing 99.671% availability.
Tier II
Tier II is composed of a single path for power and cooling distribution, with redundant components, providing 99.741%
availability.
Tier III
Tier III is composed of multiple active power and cooling distribution paths, but only one path active, has redundant
components, and is Concurrently Maintainable, providing 99.982% availability.
Ready for a little real world translation?
If you assume that the average system outage lasts four hours...
|
Tier
|
% uptime
|
period between failures (est)
|
|
Tier I
|
99.671%
|
1.6 Months
|
|
Tier II
|
99.741%
|
2.1 Months
|
|
Tier III
|
99.982%
|
2.5 Years
|
|
(ref)
|
99.999%
|
45.7 Years
|
formula: (365 days/year) * (24 hours/day) * (60 minutes/hour) * (1 - .99671) yields 2.5 years between
failures. This assumes the average failure requires 4 hours recovery.
|
|
atac data center, duffield virginia
