As they gaze at their data centres and their 2015 budgets, CIOs face an array of conflicts: Expand on-premise servers and storage, or go to the cloud? Public cloud or private cloud? How much to put into the cloud? If a business unit is successful will that mean service demand will go up exponentially? What about the installation of new equipment now — will it be done on time?
According to the authors of a recently-published book on running clouds, it boils down to two questions: What will they need to buy in the coming year, and when will they need to buy it?
That’s according to an excerpt from The Practice of Cloud System Administration, Volume 2 by Thomas A. Limoncelli, Strata R. Chalup and Christina J. Hogan, published this week on Computerworld U.S.
Some of the chapter is pretty basic: All corporate resources need to be inventoried and tracked, from operating systems to storage devices. Still, I was struck by two formulas the authors say administrators can use after finding out which components can influence service capacity, expected and planned growth rates.
Future Resources = Current Usage x (1 + Normal Growth + Planned Growth) + Headroom
You can then calculate for each resource the additional capacity that you need to purchase:
Additional Resources = Future Resources ñ Current Resources.
It sound simple, but, the authors caution, this has to be done for each standalone and shared resource.
Is it harder than reasoning, “We’ve grown three per cent for the past two years and I guess it’ll be the same next year”? Yes. On the other hand, there’s a risk relying strictly on the past will get you into trouble.