We discussed how redundancy works and what types are possible in Part One. Now let’s apply that knowledge to IT infrastructure.
2N Redundancy – The Gold Standard
2N redundancy (and above) is the gold standard, but it is also expensive and wasteful. At least half your costly equipment is essentially idle in a 2N system. Many businesses also have a hard time making space for one data center, much less two. 2N makes sense for large operations that will lose huge amounts of money during a system outage and subsequent downtime, such as e-commerce, content delivery, colocation, finance and manufacturing. It also makes sense for applications where a system failure will endanger human life, such as healthcare, public safety, transportation and defense.
N+1 Redundancy – Making Smart Choices
What if you can’t justify or afford the expense and inconvenience of having a duplicate data center waiting in reserve? Can redundancy still help you? Yes—you will benefit from a selective approach to redundancy like N+1.
Identify weak points in your system where the failure of a single critical component will bring down the whole system (single points of failure) and provide redundancy in those locations—especially where a component is more prone to failure because it is inherently unreliable or subject to wear and tear.
Example – Call Center
Your call center operation has to be operational around the clock to meet the terms of your service level agreement (SLA) with clients, but you can’t afford the capital expense of building a duplicate on‑site data center. However, you have identified two primary culprits in any outage scenario: a loss of power and a loss of internet connectivity. UPS systems will provide backup power for a basic level of power redundancy. Uptime can be further protected by installing redundant UPS systems and independent utility power feeds, as well as dual independent internet connections and console servers to provide fail-safe remote access. You haven’t duplicated every system, but you have addressed the “low-hanging fruit” and greatly reduced the chance of call center downtime.
If you decide that a duplicate data center is a must for your call center, installing one or more micro data centers in a spare room of your building can reduce cost, disruption and deployment time. You can also size the micro data centers to match your needs, whether duplicating the entire capacity of the primary data center or some percentage that’s acceptable for temporary use until problems have been fixed.
Example – Server Cluster
Providing high-density server racks with redundant power becomes more difficult as wattages increase. A 2N system with 3-phase A/B power feeds to dual PDUs and dual server power supplies is very reliable, but also costly. Leveraging the inherent resilience and automated failover of the server cluster will allow you to design power redundancy that reduces both capital expenditures and operating costs, potentially saving thousands of dollars per rack. A 3‑phase ATS can connect dual 3-phase power feeds to a single PDU and a single power supply per server, delivering high reliability and high efficiency at greatly reduced cost.
Cooling is another area that can benefit from a nuanced approach to redundancy. Providing 2N or even N+1 redundancy for computer room air conditioners (CRACs) is expensive. You can reduce costs by providing partial redundancy (think of it as N+½) through close-coupled cooling units, such as portable air conditioners and in-row air conditioners, allows servers to continue operating until CRAC repair/replacement is complete. Most IT equipment can operate at higher temperatures for hours or days without significantly reducing system availability or equipment lifespan. The key is determining costs and tolerances, then balancing them to achieve your goals.
How can Tripp Lite help?
Tripp Lite offers a large selection of cost-effective IT infrastructure solutions compatible with various levels of redundancy:
- UPS systems, including models that support redundancy.
- Power distribution units (PDUs) including ATS models that provide redundant power to single-cord devices.
- Cooling solutions perfect for providing redundant close-coupled cooling or partial redundancy for perimeter systems.
- Connect redundant systems and power supplies with copper network cables, fiber network cables and power cords.
- Micro data centers enable redundancy at lower cost than building dedicated IT facilities, with easier setup.
Better yet, Tripp Lite will help you analyze your data center and incorporate redundancy efficiently, in a way that balances your uptime goals with budget limitations. Contact us for a free IT infrastructure assessment today!
Tripp Lite will help you analyze your data center and incorporate redundancy efficiently. Contact us for a free IT infrastructure assessment today!
David Zomaya is a technical writer and business/quality analyst from Chicago, Illinois. He has a professional background in software, networking, power electronics, customer service and tech marketing.