How I decided to go to the cloud

I don’t hate the cloud — seriously. In fact, ClipTraining, the company I co-founded, uses cloud-based infrastructure to deliver its video training to clients. Using cloud-based infrastructure keeps us worry-free around availability and scalability. We weighed the pros and cons, then decided it makes sense for us.

That’s the important take-away: You must consider the good and bad to determine if the cloud is where your data center infrastructure should reside. There’s no one single formula. Plus, with all the hype, it’s easy to miss some of the pluses and minuses.

[ For the full scoop on the state of the cloud in the enterprise, check out InfoWorld’s Cloud Computing Deep Dive PDF special report. | Stay abreast of key Microsoft technologies in our Technology: Microsoft newsletter. ]

Scalability. Scalability is one of the great things about cloud-based infrastructure. If you need to scale quickly, you can do so without shopping for more hardware, deciding on virtualization products, and worrying about finding the right people to support your conventional IT infrastructure. Of course, in companies that have an infrastructure in place, that’s not to say you can’t find the space or power to spin up additional VMs when you need them — you can. Bureaucracy may be the holdup internally, but let’s face it: That same bureaucracy would likely get in the way of commissioning more cloud resources, too.

Availability. Availability is also a blessing, but not as pure an asset. I have experienced brief outages, but they didn’t cause major problems among my clientele, who look to us for on-demand video training, which typically lacks the urgency of email or banking. However, some very notable disruptions at Amazon.com over the last several years have taken down popular services for one or more days. You need to factor the implications of possible outages on where and whether you rely on cloud infrastructure.

Security. You might have noticed that I didn’t mention security as an item my company worries about. Certainly, security issues may arise, but our cloud-stored data isn’t high-value. However, we would worry if we stored other kinds of data in the cloud. Last year, for example, Amazon-owned Zappos was breached, and credit card data was possibly lost. What does this prove? That cloud-based solutions are vulnerable, just like your on-premises systems might be.

The difference is that I don’t know who’s responsible for the security policies being maintained with my cloud infrastructure. I don’t know how safe my data is, and I cannot check personally to ensure policy is being followed whereas I can if it is on-premises. If security is a big concern — as it would be for hospitals, banks, and government intelligence agencies — you might want to keep that data behind your own firewalls, handled by people you reviewed and hired personally.

Cost. You’ll need heavy accounting smarts to work through this area. Many IT admins prefer to go with cloud services because they are overwhelmed with the stress of day-to-day operations, often due to understaffing. They don’t have time to handpick every last piece of hardware. Plus, when they compare the basic cost of storage in the cloud versus on-premises, the cloud appears more attractive.

But remember that data has to be accessed and changed, read and written to, over and over again. The cloud’s costs grow with your level of access, so you need crafty accounting work to determine if the cloud will save you money versus owning your own infrastructure. Often, owning beats renting over the long term. But each organization is unique and needs to make those calculations itself.

Perhaps your organization has a need for cloud-bursting, where you access high-performance computing at a moment’s notice periodically. Being able to spin up systems quickly and pool them together is a significant advantage of the cloud in such cases; you’re not saddled with all that gear (and its support staff) sitting idle the rest of the time. But the frequency and intensity of your spikes will determine whether paying for it on demand via the cloud or paying for it in-house is cheaper — it’s not a given either way.

It’s about balance Although startups are held up as all-cloud paragons, they’re not close to the typical reality. Most companies will have a mix of cloud and on-premises resources. For them, the issue is achieving the right balance.

For example, the gaming company Zynga (makers of Farmville) use both on-premises systems and the cloud-based Amazon Web Services (AWS). Their balance has changed over time; due to availability issues at AWS, Zynga has shifted more heavily to on-premises resources, with AWS now accounting for 20 percent of its IT capability. But going all-on-premises is as senseless for Zynga as going all-cloud would be. Each company needs to figure out is right ratio and adjust it periodically as needs and realities change.

I’ve worked quite a bit with cloud-based infrastructure as a service, and I do love it — when it fits my needs. But I don’t recommend it as the magic bullet that cures all that ails you. Weigh your options, and seek a balance in what you choose.

This story, “How I decided to go to the cloud,” was originally published at InfoWorld.com. Read more of J. Peter Bruzzese’s Enterprise Windows blog and follow the latest developments in Windows at InfoWorld.com. For the latest business technology news, follow InfoWorld.com on Twitter.