Are ‘private clouds’ evil? Or just vendor hype?

There has been a great deal of back in forth in the blogosphere around the use of private clouds, which generally means IT infrastructure that relies on the same techniques cloud providers use for their own datacenters, such as multitenancy, virtualization, Web delivery, and highly standardized environments. Most of the discussion emerged from a blog post by Appirio, “2009 prediction: Rise and fall of the private cloud.”

The following comment is what set off the firestorm: “Here’s the rub: Private clouds are just an expensive datacenter with a fancy name. We predict that 2009 will represent the rise and fall of this overhyped concept. Of course, virtualization, service-oriented architectures, and open standards are all great things for every company operating a datacenter to consider. But all this talk about ‘private clouds’ is a distraction from the real news: The vast majority of companies shouldn’t need to worry about operating any sort of datacenter anymore, cloud-like or not.”

Of course there were some valid responses supporting the private cloud concept, including this post by beaker: “If we’re talking about a large, heavily regulated enterprise (pick your industry/vertical) with sunk costs and the desire/need to leverage the investment they’ve made in the consolidation, virtualization and enterprise modernization of their global datacenter footprints and take it to the next level…”

[ Stay up on the cloud with InfoWorld’s Cloud Computing Report newsletter. | Confused by the cloud hype? Read InfoWorld’s “What cloud computing really means” and watch our cloud computing InfoClipz. ]

Fact is, the term “private cloud” is a legitimate architectural pattern and option that has value within some problem domains. Typically, private clouds are virtualized architectures that exist within private datacenters but vary greatly in configuration and technology. There is no standard for what constitutes a private cloud.

I did have some concerns in calling such an architecture a “private cloud” — though I realize that doing so has the advantage of riding the hype around cloud computing — but indeed, they share some patterns with public cloud offerings. For example, private clouds can share resources more efficiently and provide better elasticity, considering that applications and data are not bound to a single server using a virtualized architecture. Thus, I’m OK with the term. But whether you need a private cloud is another story.

The issue around private clouds is not around the definition. The issue is around the marketing hype behind private clouds and perhaps the misuse of the private cloud concept as a “replacement” or “prerequisite” for public cloud computing, SOA, or good architecture.

Much of this private cloud hype has been driven by the larger hardware and software vendors, who saw early on that public cloud computing was a threat to their market, and quickly focused their marketing efforts around selling so-called private clouds — old wine in a new bottle. Therefore, they’ve tried to convince IT to reinvent the datacenter as if they were an internal cloud provider to their business, as a way to prepare for eventual moves to the public cloud (when — wink, wink — those public providers finally get the cloud secure, reliable, and compliance-capable). Sorry, guys.

The fact of the matter is that your requirements should drive you to the correct solution: private, public, hybrid, or perhaps no clouds at all. In many instances, public clouds are the most cost-effective solutions, and you don’t need to stand up private clouds first — no matter what your hardware and software vendor says. However, in some instances, private clouds may be a better fit — but at a greatly reduced cost/benefit ratio compared to public clouds. Typically, if the cloud deployment model makes sense for your IT delivery, private clouds are a fit if security and privacy is a concern and/or you already made an investment in the hardware and software resources, so you might as well use them.

Private clouds are not evil, but they could be overhyped and, at the end of the day, not all that important for many companies.