There is a sea of data center design guidelines and standards that IT leaders must wade through to select, build and implement the most effective and efficient data center. This monumental task is only compounded as IT leaders must also weigh risks to the business, such as downtime, plan for the initial capital cost and manage ongoing operational costs of the data center.
This year alone, more than half of businesses will modernize their data centers, according to Analytics Insights. By 2020, the non-stop release of next-generation technologies will require upgrades and faster adoption of new data center design innovations, such as new data center infrastructure design. Here are some foundational elements to keep in mind while selecting the right data center design for cost savings.
Begin with the End in Mind
Data centers are generally defined by the level of available capacity. While the level of capacity determines data center design, it also determines the initial capital cost and ongoing operating costs of the data center. As a result, choosing the capacity level of a data center is a preliminary and critical task.
In order to select the best capacity level, IT leaders need to evaluate at least three important aspects for the data center design: 1. Risk of downtime; 2. Impact of downtime to the business; and 3. The incremental costs of improving availability.
Because a single data center standard is not always enough to cover the scope of a business’ requirements, IT leaders may need to consider combining two or more options. In fact, using more than one data center facility may be an effective way to boost capacity. For example, building two data centers with lower capacity levels (for example, TIA-942 Rated-2 or Uptime Institute Tier II) and running them in active/sync mode (data continuously synchronized across sites) might prove a better alternative than a single data center with a higher capacity level (for example, TIA-942 Rated-3 or Uptime Institute Tier III). This is because a localized disaster can render a single data center out of commission.
Another factor to consider early on is load-balancing. Client-facing or Web-based workloads in multiple sites can often create a virtual “Tier IV” environment—at least from the perspective of the user—as workloads dynamically shift to the active site should one site fail. From a reputation management and customer service perspective, this can prove critical when the perception of 100% availability is desired.
Finally, IT leaders may need to think through building differing levels of capacity into specific zones within the data center. While redundant power and telecom feeds to the building may be required to do this, providing fully redundant active feeds to all IT equipment may not. By just focusing high levels of redundancy on mission-critical applications (and compute zones), overall construction and operational costs can be reduced while still retaining the high-capacity goals for core business systems.
Navigating Design Landscape
Data center availability guidelines and design standards have formally existed since the early 1990s. Since then, the data center industry has significantly expanded to include hundreds of new guidelines, standards, regulations, and entities. Today, there are three data center design standards, which are the most adopted standards in this industry:
- TIA-942 from Telecommunications Industry Association (TIA) in conjunction with the American National Standards Institute (ANSI).
- BICSI 002 from Building Industry Consulting Service International in conjunction with the American National Standards Institute (ANSI).
- Uptime Institute a division of The 451 Group.
While these most-accepted standards have some differences, ultimately, they share commonalities in Tier I to Tier IV requirements and ratings for data centers. Here is a breakdown of the criteria:
- Data Center Topology – Fault Tolerant;
- Downtime Hours Per Year – 0.4;
- Availability Percentage – 99.995%; and
- Additional Characteristics –
- Provides site infrastructure capacity and capability to permit any planned activity without disruption to the critical load
- Fault-tolerant functionality also provides the ability of the site infrastructure to sustain at least one worst-case unplanned failure or event, with no critical load impact
- Requires simultaneous active distribution paths.
Additionally, there are many other widely-used data center guidelines, standards, and codes that IT leaders should research and implement such as:
- American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE)
- Leadership in Energy and Environmental Design (LEED) controlled by the U.S. Green Building Council (USGBC)
- International Data Center Authority (IDCA)
- Specific standards, such as those for cabling and fire protection.
- Nonvoluntary local codes, such as the National Electrical Code (NEC), among others
Layering Capacity Levels to Save Money
The first priority for most data center guidelines and design standards is capacity—as opposed to cost. However, IT leaders can apply several best practices to significantly reduce the initial capital cost as well as the ongoing operational cost of the data center.
IT leaders will need to evaluate data center costs per capacity level, looking at both the construction costs of data centers at a given rated/tier/class level and the relative costs of attaining higher levels of capacity (essentially the cost-benefit of moving from rated/tier to rated/tier).
For example, evaluate the average costs of building an 8,000 square-foot data center at differing rated/tier/class levels, then break down the data center into three power density zones:
1. 15 kilowatts (kW) per rack for high density (15% of the data center floor space)
2. 10 kW per rack for medium density (20% of the data center floor space)
3. 5 kW per rack for the remainder of the floor space
This discipline will provide perspective and allow for options based on different variables.
Deciding Shades of Green
The expectation of environmentally-friendly data centers does not simply rest with using the newest green technology. Instead, it is about using technology efficiently to reduce operating costs in the long term, with the value-added benefit of becoming more environmentally friendly. This becomes the win-win situation where IT reduces day-to-day operating expenses, the business’ corporate sustainability ranking improves, and the result helps the environment.
There are many data center design standards available in this area; however, a LEED-certified data center typically has the following characteristics:
- Advanced cooling system to reduce energy consumption
- Improved cooling efficiency
- Reduced energy consumption
- A clean backup power system
- Renewable energy
- Green construction
IT leaders can factor in some green design elements that will make a big difference in the short and long run up to receiving full LEED certification of the data center.
At Aligned, our mission is to make critical infrastructure smart enough to continuously improve both its economic performance and environmental impact, delivering a noticeable business advantage. Sustainability helps ensure business continuity by conserving resources. By using a fraction of the space, electricity, and water of traditional technologies, we deliver tremendous results to our clients.
We are harnessing the power of intelligent infrastructure—where our collective technology, methods, and ideas can be a catalyst for that change. By helping create environments that are dynamic, ever-improving ecosystems, and by treating infrastructure like the critical resource it is, we are helping to improve your bottom line. To learn more, please contact us today!
