Multiple Views of the Private Cloud


Ever since the notion of cloud computing started to take hold, the idea of leveraging the same technology behind the enterprise firewall has proven to be equally appealing.


These private clouds would be capable of shifting enterprise data loads across increasingly dynamic infrastructures, at once lowering costs by utilizing idle or underused network elements and increasing flexibility with the ability to allocate resources where they are needed most.


Not everyone is sold on the concept, however. InformationWeek's Andrew Conry-Murray raised some dust earlier this month with a blog arguing that the term "private cloud" is a misnomer at best because one of the core concepts of cloud computing is that practically the entire infrastructure is provided by a third party. His take is that dynamic networking may be a worthy goal, but it has no real connection to the cloud.


InfoWorld's Eric Knorr carried the ball a little further, suggesting that, regardless of what you call it, the promise of an uber-flexible enterprise infrastructure that can provide all things to all users has been an IT chimera for years. He's not holding his breath for it anymore.


Countering those arguments is Cisco's James Urquhart, who says that the private cloud is nothing more than the continuation of numerous Internet technologies that either originated or were duplicated in the enterprise, the primary example being the intranet. And since a private cloud will offer a substantial improvement over current silo architectures, enterprises will embrace the concept even if it doesn't prove to be the ultimate IT fantasy solution.


Even so, the ability to launch and manage a private cloud will most certainly be reserved for the largest of the large, says John Foley, again on InformationWeek. Since the major portion of the private cloud's ROI is derived from energy savings and lower personnel costs, you'd have to have a pretty substantial infrastructure already in place to make it worth the investment, which can run as high as $2,500 per server. Small and medium-sized firms would do better on a public cloud.


Semantic arguments notwithstanding, I agree with Urquhart on this one. Of course the private cloud won't fulfill all the dreams of its strongest proponents. But neither will the public cloud. No technology ever does. But there's enough to the private cloud concept for large firms to give it a go, if only to find out if the concept can hold up under cold, hard reality.