The uptime
institute has established four levels of fault tolerance for data centers. The
below tier classification has been an accepted standard for defining the levels
of the data center fault tolerance. A data center is said to be concurrent
maintainable, only if they are able to perform the planned maintenance activity
without shutting down the critical loads.
TIER 1
|
Single
path for power and cooling distribution; no redundant components – less than
28.8 hours downtime/year
|
TIER 2
|
Single
path for power and cooling distribution; Redundant components – less than
22.0 hours downtime/year
|
TIER 3
|
Multiple
power and cooling distribution paths, but only one path active; redundant
components; concurrently maintainable – less than 1.6 hours of downtime/year
|
TIER 4
|
Multiple
active power and cooling distribution paths; redundant components; fault
tolerant – less than 0.4 hour of downtime/year
|
Source: The Uptime Institute
·
TIER 1 Data Center
A TIER 1
data center is a basic data center, which was first deployed in 1965. It is
prone to disruptions from both planned and unplanned activity. This type or data
center may or may not have a generator or UPS. In order to perform annual
maintenance or repair work, the data center should be completely shut down.
·
TIER 2 Data Center
A TIER 2
data center has all the features of TIER 1 which was first deployed in 1970. It
has redundant components and less prone to disruptions from both planned and
unplanned activity. This data center has a raised floor height of 18 inches and
has multiple servers, UPS and generators. The redundant component is “Need plus
one” (N+1), which is a single threaded distribution. In this the applications
are covered by some kind of Business continuity or Disaster Recovery plan.
·
TIER 3 Data Center
TIER 3 data
center was first deployed in 1985, which incorporates all the features of TIER
2. In this tier, planned maintenance activity (Like repair and replacement of
components, testing of components, programmable maintenance, etc.) can be
performed without disrupting the computer hardware components. But the
unplanned activities (Like failures in facility infrastructure components) will
result in disruption. This type of data center has redundant power, cooling and
networking systems. It is highly secured and can handle up to 72 hours of power
outage.
·
TIER 4 Data Center
A TIER 4
data center has all the features of TIER 3 and is fault tolerant. This was
first deployed in the year 1995. This type of data center has active
distribution paths which is System + System configuration (i.e. it has two
separate UPS system with N+1 redundancy in each system). In this tier, all the
computer hardware requires dual power inputs. The rooms and zones are isolated
and has the raised floor height of 30 to 36 inches. In order to achieve high
availability, reliability and serviceability, this tier employs Clustering,
Direct Access Storage Devices (DASD), 24/7 monitoring, Thermal storage and many
more. Tier 4 data centers are the most expensive to build and maintain.