As with all things "cloud" these days, there's an interesting dichotomy between those who are head over heels excited about the huge potential of private cloud approaches and those who see it as a marketing-driven rebrand of what everyone is already doing: virtualizing the data center. In the end, the truth is somewhere in between.
Although the private cloud has the potential to solve many of the scalability and service-delivery challenges that larger IT departments wrestle with, it is not magic. Really reaping the benefits of the private cloud requires an IT-savvy customer base, a willingness from the executive suite to adopt a chargeback model, and no small amount of work to configure it in such a way that it will actually save everyone time in the end.
At the same time, a true private cloud implementation is not just data-center virtualization with a new coat of paint. Although it's true that much of the gear you might use in a private cloud ends up being put together in the same way as you would in a traditional virtualization model, the self-service nature of the private cloud demands a completely different mindset from those planning and building the underlying virtualization and storage infrastructures.
[ In the data center today, the action is in the private cloud. InfoWorld's experts take you through what you need to know to do it right in our "Private Cloud Deep Dive" PDF special report. | Also check out our "Cloud Security Deep Dive," our "Cloud Storage Deep Dive," and our "Cloud Services Deep Dive." ]
What's different about private cloud storage
In a typical IT department, planning and budgeting for new infrastructure expansion usually happens many months before the beginning of the fiscal year and is based on the perceived demand from business units for various data center resources -- whether they be compute, network, or storage. However, even though that process is becoming harder and harder to do accurately -- especially as storage demand grows at ever-increasing rates and becomes more difficult to predict -- that's still the way most of us do it.
In the end, IT is ultimately in the driver's seat: deciding what infrastructure expansions are made and when. When a business unit inevitably discovers that it needs to dramatically increase its resource consumption (say, to field a new application), a usually lengthy process commences of figuring out where the money will come from. Finally, after a few months of wrangling, the service is delivered and life goes on.
Therein lies one of the key differences between a traditional data center and a private cloud infrastructure. Due to the self-service nature of the private cloud, you may have literally no warning whatsoever when an enormous usage spike might occur. Thus, you need to maintain enough slack capacity to absorb a reasonable amount of growth and be able to react extremely quickly to satisfy demands that outstrip your available capacity -- much in the same way that a typical service provider needs to.
Staying ahead of that kind of unpredictable growth demands an infrastructure that is designed to be grown in small increments under very short timetables, as well as the implementation of monitoring tools that can track and trend the consumption rate of available resources.