Decoding data center redundancies, and why component quality matters

In the English language, redundancies would refer to superfluous repetition and is really something that people tend to want to avoid either in speech or writing. In engineering though, and especially in data centers, redundancies are something you want to have. At a data center, redundancy refers to a system design where a component is duplicated or multiplicated so that in the event of a component failure, equipment and processes are not impacted.

Here, redundancies are all you can count on in the worst-case scenario, like a natural disaster or a power outage.

In Uptime Institute’s 2021 Global Data Center Survey, over 60% of the respondents reported losing more than $100,000 to downtime, and of that 60%, 15% lost over $1 million. With more enterprises moving to the cloud, the maximum tolerable period of disruption is decreasing as there is less tolerance for operations experiencing any sort of downtime. Your bottom line, business operations, and customer experience could hinge on how quickly your data center can recover from a disruption.

Data center redundancy is vital to ensure zero downtime.

There are different kinds of redundancies in use at data centers, the most common being N, 2N, and N+1. N refers to the bare minimum number of resources required to operate an IT system. N+1 would mean there is an extra duplicated resource, and 2N would mean that there is twice the amount of required resources/capacity available in the system.

Decoding the ‘N’s of redundancies

Take the case of a server in a data center that has five servers with an additional five that act as a dedicated capacity. This reserves a combination of 20 servers in total, which is a 2N redundancy.

If you do not have redundancy, when there is a failure, the system will be non-functional and inaccessible until the issue is resolved.

N+1 would signify a single backup. While this redundancy level may work for a small IT system, it is risky for medium- to large-scale enterprises.

N+2 would be the next best thing to N+1. It consists of the resources needed for the system functionality plus two separate backups.

You can keep adding resources in this fashion, still, most small enterprises stick to N+1 and N+2 unless there is a specific or unique requirement to keep multiple redundancies. With the proliferation of dual power supply IT infrastructure, larger enterprises especially healthcare facilities prefer 2(N+1), which means twice the required resources/capacity to function normally plus a backup.

One may assume, more is safer, but remember your redundancy level depends on:

How to arrive at your ‘N’ formula

• Budget

• Business goals

• IT Environment

• Compliance requirements

A look at the Ns would tell you that redundancies depend greatly on the reliability and quality of individual components. If individual components sourced have low failure rates, it will be easier to mitigate any issues that arise.

At Hardy Racks, we offer a range of data center accessories that have only one thing in common – high quality. Whether power distribution units or server racks, fiber raceways, or temperature control units, our products offer the resiliency and dependability you seek.

Contact us today to learn more.