Top Considerations When Defining Your Data Center Needs

Data Centers
Jul 1 2015

Remember: No two data centers are built the same. Each solution has relative strengths and weaknesses and

Mature businessmen deep in discussion while using a tablet

Mature businessmen deep in discussion while using a tablet

finding the perfect solution for you makes thorough planning imperative.

Defining the resource needs 
There are core areas where a data center facility must be flexible in providing the right type of infrastructure support.  In working with a data center provider, ensure that there is room to grow and that the environment is capable of handling the needs of your organization. During the planning and early development stages, data center and other technology leaders should focus on the following:

  • Space – This is usually defined by floor space as well as rack space. Furthermore, if a distributed environment is needed, ensure that your provider is able to accommodate future plans for growth.
  • Power – Numerous variables fall into this planning metric. Working with a partner that can support a range of densities and power distribution models is extremely important. This means that both current and future equipment can be supported. High-density computing, although improving in efficiency, can consume more power than standard rack-mount servers.
  • Cooling – There are so many new types of architectures being deployed that each rack could have very unique systems that it’s supporting. Controlling room and aisle environmental variables is important – but gaining control within the rack itself is critical. Modern rack airflow management products are specifically designed to control intake airflow in server racks. This is critical because your rack could have variable cooling requirements. Consider this – modular airflow containers are being deployed within next-generation data centers and are specifically designed to block airflow, ensuring hot and cold aisle separation. Capable of controlling a variety of data center configurations, delivering rack-top vertical baffles, and even bi-directional containment; these kinds of cooling solutions allow your data center to be a lot more agile.

When evaluating different types of cooling solutions, take advantage of the experience of data center cooling system experts from the Data Center Group of Nortek Air Solutions. They have 30 years of experience providing cooling infrastructure for mission critical facilities and can help you determine what type of system your data center needs. As a custom manufacturer, Nortek Air Solutions can design and build to exact performance requirements such as efficiency, footprint, sound and configuration.

  • Redundancy – When defining your needs for electrical redundancy, you should realize that “2N” is not always “2N.” To truly define redundancy, it should be from the electrical grid or primary source through the “last foot” of electrical distribution. For example, dual corded server connections do not – on their own – imply “2N” as there may be single points of failure in other components of the system. Ensure that the provider is able to offer N+1 or higher redundancy on separate, physically provided, circuits. The amount of redundancy offered will impact cost structures. Therefore, it’s important to thoroughly understand the actual amount of redundancy required by your organization – and how that will be delivered by the providers.

Cooling system redundancy is also important. One example of this redundancy is the FANWALL® fan array system from Nortek Air Solutions. With conventional single fan and motor systems, a failure of either component shuts down the entire air handler system.  This creates a critical path failure that requires immediate, costly action to restore function and minimize revenue and profit losses.  When a fan or motor failure occurs in a FANWALL system, the remaining operating fans can compensate to maintain airflow until the failed component(s) can be repaired or replaced.

  • Utilization – The planning stages will analyze how the environment will be utilized both now and in the long-term. This can help define critical systems and allow your data center provider to ensure you have the right solution in place. If some systems require more care than others, look for a partner which can adapt to those types of needs.

Having the end (goals) in mind
The challenge with data center design is that technology can rapidly change. In many situations, users don’t, or can’t, accurately predict their 5 to 10-year growth rate. When working with a data center provider, the ability to scale and be agile in conjunction with the organization will be crucial. This is why a seemingly attractive long-term solution can prove to be very costly if scalability isn’t built in.

Analyzing the “current vs. future” scenarios including IT refresh cycle
The refresh cycle of any organization should be a serious consideration point when defining a data center solution. In some cases, this can be built into or around a contract. By understanding both current and future demands, data center managers can work with providers and ensure that they scale when needed. Without a good idea of when hardware reaches end-of-life, data center technologies can become old and dated. This could result in your organization being stuck with an inflexible data center lease that does not support evolving hardware densities, rack configurations or network connectivity.  Initial capital investment in networking cabling and support infrastructure can also prove very costly to abandon in place and relocate to a different data center if needs in the existing facility exceed capacity of any of the above listed factors.

There are a lot of new technologies being developed which can better align with your data center needs.

New concepts around power and utilization management are re-defining data center environmental controls. Now, there are solutions around advanced airflow management and even “free cooling.” One of the newest free cooling solutions is the Cool3 ™ indirect evaporative cooling (IDEC) unit from Nortek Air Solutions which offers maximum water efficiency and ultra-energy efficiency with typical pPUE below 1.1. These kinds of solutions are critical to consider. Why?

  • Free cooling can offer impressive energy cost savings for data centers
  • New recommendations for both thermal and dew point ranges mean free cooling can be used in more climates
  • Several available methods of free cooling provide data center operators with the opportunity to select the option that best meets their specific system requirements.

Once the above needs are outlined, data center managers can really begin to identify the right type of partner and data center design. In some cases, working with a pre-defined data center is the right move. However, in many other situations, data center managers actively strive to build a data center environment they can call their own—an environment designed around efficiency, control, and a good business use-case.
For more information about innovative cooling systems for your data center, or to put one of our Data Center Group experts on your facility planning team, contact us at datacenterinfo@nortek.com.