Designing and building a new data center that aims to be sustainable and efficient – particularly without having to compromise on availability (uptime) performance – is a challenging process fraught with potential pitfalls. In your data center design planning, not only do you have to make correct decisions, but you have to make them in the proper order.
The repercussions of poor design planning can include not only wasted time and effort and costly re-design work, but worse, they can cause you to be late to market if you’re a service provider and/or end up with a facility that isn’t competitive or doesn’t meet your business needs. If you’re an enterprise, this might force you to re-evaluate your outsourcing strategy in terms of having to find more collocation and cloud services that do meet your needs.
To help you successfully navigate the design process, here are three common mistakes that companies make in the data center planning and design stages and how you can avoid them:
Mistake #1: Improperly setting design criteria and performance characteristics
It’s only natural to want the best performance, but unless you have the resources of an Amazon or Microsoft, you probably don’t need a Tier IV data center; in other words, don’t overbuild. It ends up costing more upfront, reduces electrical efficiency and will be more expensive to maintain and operate the facility over the long run. Non-technical project stakeholders, in particular, might ask for the best without fully understanding the consequences on things like cost and time to market. Fully realizing the impact of these early decisions might come too late in the design or build phase, long after a lot of investment has already been made.
Schneider Electric has developed several free tools for people to use in the early planning phase of data center projects. Data Center TradeOff Tools make it very easy to quickly compare and contrast various design alternatives so that even non-technical project participants can understand the tradeoffs of various design decisions. These tools can save you time and energy. Our Data Center Reference Designs are another great free tool for helping you move through the early planning phase quicker and through the detailed design phase with fewer errors. Our library of reference designs offers nearly 100 high level conceptual plans for how a data center is built. Each design covers the electrical, mechanical and IT space systems for a given set of performance characteristics.
Another mistake companies make is getting too hung up on the minutiae of technical specs, the ‘speeds and feeds’ that don’t necessarily align with the company’s business goals and risk profile. Keep your eye on the big picture. It is critical to focus on the big, key project parameters of criticality (ie, Tier level), efficiency, budget, capacity, power density and the future growth plan. These parameters pretty much drive everything else about the design. White paper 142, Data Center Projects: System Planning goes into each of these in detail.
Planning failures, wasted design work and schedule slips are often traceable to not gathering all the input with key stakeholders on the front end. Data center project managers need to forge a shared understanding and agreement about these key project parameters.
Don’t forget to include leaders from finance, IT and business at this early stage in the process. Conduct workshops in which those tradeoffs can be discussed and agreed upon.
Mistake #2: Site selection before confirming data center design criteria
A common “cart before the horse” mistake is to waste time and energy searching for a site location before the data center design criteria have been nailed down. Deciding on a location or specific geographic area without understanding and agreeing to the full scope for the data center can lead to problems.
When selecting a site prematurely or based on narrow geography, you could jeopardize what site design requirements are even possible. For instance, you could find limitations with things like availability, reliability and costs associated with the electrical power grid, the water supply and fiberoptic cable. Other site criteria include proximity to emergency services like fire stations, or to transportation features like major highways.
Data centers should also be located with disaster avoidance and preparedness in mind – companies should be aware of the potential risks associated with flash floods, tornadoes, hurricanes, earthquakes or other natural disasters.
On the flip side, building a new data center can also provide an opportunity to create a green, sustainable, environmentally friendly facility. Therefore, the design criteria and site selection should also take into consideration opportunities for generating energy onsite from renewable sources, such as solar or wind power, either by yourself or through an independent power producer (IPP). Utility providers, power purchase agreement terms and the existence (and cost) of renewable energy credits and/or carbon offsets are other key site considerations when energy and sustainability goals are important. For more information on criteria for data center site selection, see White Paper 81, Site Selection for Mission Critical Facilities.
Mistake #3: Space planning before design criteria is firmed up
Some organizations base their site search criteria on the amount of raised floor required to house their critical IT infrastructure, which can be significant. In the most robust of systems, the ratio of IT floor space to support gear could be as high as 1 to 1. But this can lead to a major mistake.
Mechanical and electrical equipment require a significant amount of space. In addition, many organizations overlook the square footage required to house office space, equipment yards and IT equipment staging areas. Therefore, it is absolutely critical to determine your design criteria before you develop your space plan. Without it, there is no way to conceptualize the total space required to meet your overall needs.
It’s important to create a detailed data center floor plan early in the design process. Floor plans should be considered part of the preliminary specification process and should be determined before detailed design begins. See White Paper 144, Data Center Projects: Establishing a Floor Plan.
Getting your data center ducks in a row
The process of designing and building a new data center is complicated enough without making mistakes that can cost time and money. If the right issues are resolved in the right order by the right people, vague requirements can be translated into detailed design plans that meet the needs of the business. In the all-important planning phase, companies need to establish key project parameters, develop a system concept, incorporate user preferences and constraints and determine implementation requirements. For a more detailed roadmap, access White Paper 142, Data Center Projects: System Planning.
2 years ago
SE White Paper on DataCenter is great.