Redundancy and Data Centers

Redundancy, in IT Industry, in its simple description is duplicating of computer components or resources so that back up can be provided if original system fails. When it comes to data centers, Redundancy becomes one of the defining attributes for a data center and business contract negotiation. In a data center redundancy refers to duplication of system components, servers, operating systems, telecommunication links, power supplies and cooling systems that are provided for the purpose of backup, if original system fails. Any data center defines the redundancy and availability in 4 tiers: Tier 1 with 99.671 % availability (28.8 hrs/year downtime and No redundancy); Tier 2(22 hrs/year downtime and partial redundancy in power, cooling) with 99.741 % availability; Tier 3(1.6 hrs/year downtime and N+1 Fault tolerant) with 99.982 availability and Tier 4 with 99.995%(26.3 min/year downtime and 2N+1 fault tolerant) availability. (tier-standards-overview n.d.) .Tier 4 is considered most robust and reliable with lowest possibility to failures. Tier 4 is meant for business critical applications and provides fully redundant systems.

Redundancy and reliability comes with price, tier 4 being the most expensive architecture as it is least prone to failures. Many businesses fail to understand their need when it comes to redundancy; redundancy requirements for a business need to be analyzed carefully. There are certain parameters, which can help in deciding the redundancy needs for a business.

Tier 1 , least robust tier, is suitable for small businesses such as internet based startup companies, which do not have huge customer base and financial, Quality of service liabilities; small business where information technology is used for internal business processes improvements; small businesses which don’t depend on web presence as the primary marketing media.

Tier 2, partial redundant, is suitable for small business whose revenue is not directly associated with real time delivery of services such as business who can afford the downtime during off business hours ; companies which don’t have any real time delivery obligations; internet based companies which don’t have any stringent quality of service commitments; call centers businesses where more than one site is available; educational organizations.

Tier 3, reasonably redundant, is suitable for businesses that have high availability requirements for business continuity. These companies can afford the outage during an unplanned event but generally have high availability requirements. These businesses include internet based companies with stringent quality of service promises; business such as customer support ,help desks that can accept short outages; businesses which are dependent on IT resources for automated business processes and there is no direct impact of outage on the customers.

Tier 4, most reliable, is suitable for businesses whose revenue stream is directly associated with the availability. These companies have extremely high availability requirements and try to achieve maximum reliability. These businesses are generally e-commerce businesses, for instance: Amazon cannot afford downtimes because each minute of downtime is huge impact on the sales, company revenue and brand image; international companies where stakeholders access to various IT applications act as a competitive advantage; Financial institutes which support 24*7 inbound and outbound wire transfers; Some of the businesses from Tier 3 upgrade to Tier 4 over the period of time as they have understood the businesses impact from the experiences in Tier 3. (W. Pitt Turner n.d.)

Given the increasing dependency on IT resources, redundancy is becoming a necessity for the businesses and businesses are ready to make huge investments to ensure the highest availability. However, it is vital to understand the business needs and what impact the business would have in case of a failure or outage. For mid-size businesses, if the business impact is not clearly identified then try and watch approach can be followed to define the impact starting from Tier-2 and going upwards, if possible. No two businesses have exact same business impact and cost benefit matrixes. Therefore, It is very important to do the cost benefit analysis so that businesses can align to appropriate data redundancy topology and strategy.