What do Virtualization and Cloud executives think about 2012? Find out in this VMblog.com series exclusive.

Cloud Predictions for 2012

Contributed Article by Dave Laurello, president and CEO of Stratus Technologies

Half of the applications in my company run in the cloud, so when I say 2012 will be a year of "buyers' remorse" it's not because I view cloud computing negatively.

For whatever reason - executive-office pressure, a profusion of media hype, fear of being left out, protecting turf  - a hefty number of companies and IT managers will rush to cloud computing without having done proper due diligence. They will force-fit applications and portions of their legacy infrastructure into the cloud without cost-benefit analysis, a long-term strategy in place, proper vetting of prospective suppliers, or definitive metrics of success.

At year-end, too many will ask, "What have I gotten myself into?"

From our own experience with cloud computing and from working with customers, here are a few observations:

• Don't assume moving an application to the cloud is more cost-effective than running it on existing infrastructure. This is where cost/benefit analysis comes in. Understand who uses the application and how often; is the application steady state or subject to wide fluctuations or significant future growth; what other applications does it interact with; what are the implications of data interruption, data loss, or extended downtime. Once you have a profile, look at the application lifecycle in light of the terms, conditions and fees of the provider(s), against the internal investment you've made already, and your risk tolerance.
• New applications make the most sense for cloud consideration. This lets you ease into cloud computing without disrupting existing business or work practices. It gives you a chance to demonstrate success, and pave the way for more investment. It helps you to understand what it's actually like to work with a cloud service provider, test their systems, and verify their service commitments. The exception is start-up companies; go to the cloud and don't look back.
• Think long and hard before putting critical business applications in the cloud. Then, don't do it. This gets back to cost-benefit analysis. The fact of the matter is that most companies have no idea how much money it costs when critical applications go down. Of those that say they do, they grossly underestimate. The real numbers are scary. Placing critical applications in the cloud is beyond risky. It's bad business practice, especially if you have associated concerns such as regulatory compliance issues or your business reputation (for many companies, this is huge).
• The common notion is that cloud service providers offer better uptime assurance than can be achieved internally. Regrettably, this one is probably true. But it doesn't have to be that way. Most companies and all cloud service providers I've come across assume that failure is inevitable. Their attitude is, get used to it. In reality, it's inevitable that things do break, but failure is preventable. Building an infrastructure for failure recovery versus failure prevention is a complete mind-set change (like cloud computing) and a different approach to application support and infrastructure investment.

The cloud company executives I know agree that cloud computing - pardon the sports analogy - has not reached first base. It's still tail-gating in the parking lot. They also agree, as do I, that a cloud/no cloud choice is not the future of information technology. It's a combination. It's an infrastructure with an internal cloud handling critical applications and maintaining sensitive data. It's external cloud handling low-to-mid tier applications, voluminous data storage, product and application development and test. It's a hybrid cloud when some internal applications need to go external for occasional scale and then return home when the spike is over. And, the correct balance will be different for virtually every company in pursuit of the cloud's many benefits.

###

About the Author

Dave Laurello is president and CEO of Stratus Technologies, Maynard, MA. Stratus is a global company focused on providing customers with superior uptime assurance for their critical business operations and applications.