A new report released by Cushman & Wakefield and hurleypalmerflatt seems to confirm this observation. In compiling their second Data Centre Risk Index, Cushman & Wakefield and hurleypalmerflatt evaluated risk in the following 13 categories, which are then broken down by tiers (see figure 1).
· Energy – cost per kWh
· International internal bandwidth – mbps
· Ease of doing business
· Corporation tax
· Labour – cost of labour/hour
· Political stability
· Sustainability – % energy from alternatives
· Natural disasters
· Population education level
· Energy security
· GDP per capita
· Water – availability per capita
As you can tell by the spellings, the Data Centre Risk Index takes an international perspective (see figure 2), ranking the risk it sees in 30 nations.
The report is well sourced and makes good arguments for the conclusions it presents, and I’m not planning to dispute them. However, it would seem awfully difficult to address deficiencies in any of these areas once a data center site has been selected. How, for example, would an organization address high utility costs if it were located in a high-cost area?
It’s also important that an organization weight the risk categories before performing site selection. Chris Crosby, Compass Datacenters, did just that when launching his new business, which he believes will succeed by providing data center space in underserved markets. Only after determining a region and finding likely customers can Compass begin the process of determining how much space to build, what kind of connectivity will be needed, and where water, power, and fiber are available at affordable costs. Compass will also need to service its facilities, so workforce will be important too.
Once broad parameters are drawn, a data center designer and builder can wring efficiencies out of the site and optimize the building design.
It goes without saying that an organization must know a lot about itself and its business goals before knowing how the 13 factors listed in the Data Centre Risk Index affect its data center planning.