As businesses become increasingly reliant on access to digital information, the role of the data center — which houses the precious new resource known as big data — has evolved. Once considered to be necessary but expensive cost centers, data centers are now in a position to steer business results and help companies edge out competitors. But as data centers grow to become the crux upon which many businesses rely, a greater sense of urgency in being able to quickly plan, build and deploy highly efficient and cost effective data centers — with minimal disruption to business continuity — is spreading.

Unfortunately, interruptions continue to pervade both new builds and expansions.  According to a recent survey conducted by the Uptime Symposium, approximately 46% of enterprise data centers have experienced one or more business-impacting outages in the past twelve months. At a potential cost of thousands of dollars per outage, businesses simply cannot afford to depend on complicated green field or retrofit designs that require time-intensive planning and intrusive construction, and often result in complex systems that lack redundancy and predictability.

To streamline the planning process while at the same time minimizing the risk of outages (both during and after construction), data center designers and managers should consider using reference designs that offer a prescriptive and predictable approach to design, as well as standardized data center architecture concepts that promote availability through pre-assembled and integrated data center facility modules.

Planning: Utilizing Reference Designs

The data center design process isn’t an easy one and the hardest part can be just getting the project started. Poor communication within the IT department, lack of proper abstraction, tight budgets, and minimal staff are often just the first hurdles. Even after the high level design parameters (power capacity, goal PUE, etc.) are established, many challenges to quickly and efficiently deploying a data center can remain. Effective planning is key, as any mistakes made during this process can magnify and propagate through later deployment phases, resulting in delays, cost overruns, wasted time, and ultimately a compromised system.

Reference designs — tested, validated, and documented plans outlining the construction and layout of a data center’s physical infrastructure, including which components are used — can simplify and shorten the planning and implementation process by acting as a baseline design that is adapted to meet the specific constraints of the data center or preferences of the engineer. Reference designs are built upon recommended and proven best practices and because they act as a starting point for construction, can help data center managers avoid some of the most common complications associated with the design, build and expansion process, including unnecessary disruptions to business. By also offering predictable performance that can minimize risk and improve reliability of the data center, once operational, reference designs can ensure customers receive reliable service during construction and well into the future.

Beyond this, reference designs can also provide a quick and easy way for data center design teams, as well as any and all stakeholders involved in the build process, to accurately compare possible scenarios and design trade-offs. With custom design approaches, it can take days or even months for project team members to understand the impact changing variables can have on a given design. With reference designs, this process can take just minutes.

Building: Standardize, Standardize, Standardize

In today’s fast-paced business environment, it is no longer practical or cost-effective to completely engineer from scratch all aspects of a data center. Through the use of prefabricated facility modules, data center designers can realize increases in speed to deployment, predictability and reliability, agility and scalability while also efficiently managing data center lifecycle costs. In fact, when compared to traditional facilities, the use of standardized, pre-assembled and integrated data center facility power and cooling modules can be 60% faster to deploy and can provide a first cost savings of 13% or more.

Because facility modules are pre-engineered, pre-assembled and integrated, pre-tested and delivered as standardized “plug-in” modules, the unique one-time engineering, on-site assembly, installation, integration, and associated risks, such as human error, and costs common to traditional data center approaches is bypassed. Prefabrication of repeatable modules also improves compliance, safety, and efficiency as the modules are manufactured in much greater quantity than a larger non-modular system, increasing the level of quality already inherent in mass production. Additionally, in-factory design and manufacturing are closely coupled to greatly minimize uncertainty, which results in more predictable performance (read: fewer outages) of the data center infrastructure.

The use of prefabricated modules can also enable data center managers to reap the benefits of a standardized design even after the initial building process is complete. First, managers need only contract for one or two “big box” modules, rather than draw up terms and conditions for various components from numerous vendors. Additionally, because modules come to the customer already built and integrated via one manufacturer, accountability for the proper function of the module falls to the vendor, rather than the data center manager. Second, when necessary, swappable modules can be removed for factory service, enabling continuous quality improvement in which defects are diagnosed at the factory and engineered out as they are discovered. Lastly, modular design provides the advantage of fault tolerance — redundant modules operate in parallel, which means that when a single module fails, the overall system performance is unaffected, enabling business to continue to run smoothly.

As businesses continue to rely more heavily on the uptime and availability of data centers, design concepts that speed time to deployment and minimize risk in planning, building and maintenance will be necessary, and the benefits of reference designs and prefabricated modular architectures are compelling. By adopting these design concepts, businesses will be able to eliminate uncertainty, increase profitability and drive new opportunities.