Perhaps your business is growing and you need to expand your capacity to store data. Or, you’ve decided that your business has tempted fate too long without a backup data storage system. Whatever it is that triggers your decision to change your organization’s IT capability, it’s important to meticulously plan a new data center project.
A key step in planning the new (or improved) system is defining IT parameters. This involves looking at three characteristics of the IT function of the data center: criticality, capacity, and growth plan. These are the factors that will be used to determine requirements for the physical infrastructure of the data center that provides power, cooling, management, physical security and fire protection for the whole data center.
It is essential to clarify and quantify these basic parameters of IT design to further guide the data center planning sequence.
Criticality represents on a scale of 1 to 4 the importance of a data center’s operation to a business in terms of its ability to withstand downtime. The level of criticality will impact major characteristics of the system, such as redundancy of power and various details of room construction that affect reliability.
To understand the spectrum of criticality levels, consider a business whose criticality is determined to be 1, which is the lowest level. This is typically a small business with a limited online presence and low dependence on IT. Downtime is perceived as a tolerable inconvenience. A business with a criticality of level 4, the highest level, is likely a million-dollar operation whose revenues are primarily driven by electronic transactions. The business model depends entirely on IT, and downtime comes at a high price.
The design of each system would reflect this criticality level. While a level 1 business might not even have a generator, a level 4 system would be designed to include two independent utility paths, could withstand a 96-hour power outage and would include 24/7 onsite maintenance.
It’s essential to understand a business’ mission in order to determine the importance of the data center’s reliability and availability.
Ask an organization “How big a data center do you need?” and you’ve got the capacity parameter. It’s an estimate of the maximum IT power load that will be supported over the course of a data center’s lifetime.
The goal of the system capacity plan is to make sure there will always be enough power or cooling capacity to support the IT load. The capacity parameter will be used as one of four parameters in the growth plan.
Four parameters in the growth plan define the expected growth of the IT power load: the maximum final load, minimum final load and initial load (all expressed in kW) as well as the ramp-up time (years).
These four numbers make up the IT load profile used to guide the power system design. By considering maximum final load (capacity), a system can be designed to handle a company’s uncertainty about future growth. In addition, a scalable system can grow in increments over time, rather than building a system to capacity at the outset. Building a system to capacity can result in overbuilding – which can mean unused capacity and wasted capital (for unused equipment and its associated operating expenses).
Most data centers never reach their full projected capacity, but it’s essential to design a system that can support the maximum projected load if needed.
For more information on determining data center parameters, read the APC by Schneider Electric white paper, Data Center Projects: System Planning.