Editor’s note: “Cloud computing” is the hottest sector in information technology, but it’s not without risks. WRAL Tech Wire reached out to Kip Turco, senior vice president of data center operations at Raleigh-based Windstream Hosted Solutions, to explain some of the risks and choices companies must face and make in determining a cloud sttrategy. Turco is a highly sought after speaker on cloud computing topics and a frequent contributor to WRAL TechWire. He can be reached at email@example.com.
RALEIGH, N.C. – If the most recent Amazon cloud outage has taught us nothing else, it taught us that we have to keep learning. And what we learn may point the way toward the future of cloud computing.
Amazon’s Elastic Compute Cloud (EC2) suffered a three-day long outage last month. Websites like Foursquare, Reddit, Quora and others that depend on EC2 went offline, leading publications like Fortune magazine to declare the event “Amazon’s cloud nightmare.”
The headline’s inference was clear: If it happened to Amazon, it could happen to you.
Writing in The Register, Cade Metz said the event proved that Amazon’s “availability zones” “don’t exactly work as they’re designed.” And on CNET, Jonathan Eunice wrote that the outage was “giving enterprises, service providers, and developers pause–and will continue to do so for months to come.”
But once you got past the immediate “how did that happen” moment, it became clear that the vast majority of techies understood what had happened, and realized that the incident should not cause them to swear off cloud infrastructures forever. In fact, quite the opposite is true: An eWEEK reader survey revealed that 80 percent said the Amazon outage would make no difference at all to their cloud computing plans.
Writing in InformationWeek, Jonathan Feldman went one step further, calling the outage “irrelevant to those of us planning the future of computing at our organizations.” (Incidentally, Feldman is the director of IT services for the city of Asheville.)
A prime driver for this reasoning stems from the reality that, although taken for granted at times, achieving reliable performance is difficult and requires a lot of work, money, and expertise.
What the outage should do is make all of us re-examine what we expect from cloud computing and what steps we are taking to ensure that it provides us with the highest, and most continuous, level of service that we need.
Amazon’s cloud infrastructure is largely public, meaning companies can quickly and cheaply move applications to it, and gain the benefits of slashed capital expenditures and “pay-as-you-go” pricing, enabling them to ramp up or down as their needs fluctuate.
The downside of the public cloud is that companies choosing it are largely at the mercy of the platform design. And this public platform is designed for the “common” target customer that does not always have the same uptime requirements for all companies.
So, when it goes down, they have little recourse but to wait until it’s functional again and backend production support options may be limited.
Hence, there are a number of other companies for whom this model is simply not acceptable; for them, downtime is not merely an inconvenience….rather, it may mean a loss of millions of dollars, or even the future of the business itself. These firms require a far greater degree of assurance, and access. Business continuity is at the heart of their needs.
They’ll insist on private cloud implementations, featuring “five-nines” uptime in their service level agreements, and expert, 24/7 live time support, so that when something goes wrong, downtime is measured in seconds or minutes at worst, instead of hours or days.
There’s no right or wrong answer here.
Your future in cloud computing depends on your individual needs. If cost is a major factor, and you can afford the occasional outage, you may want to steer toward a public cloud implementation.
Conversely, if your company prizes reliability and consistent uptime above all, look to providers offering a “production” level private cloud product with a robust backend support structure may be a good option for you. (Interestingly, we see a trend toward hybrid cloud implementations; mission-critical applications are run in a private cloud, with less important ones sited in a public setting.)
The key here is moderation. Articles that proclaim “the sky is falling” notwithstanding, cloud computing is not going away. And yet, companies should take prudent steps, asking themselves “what if,” and what level of service they’ll truly need to ensure their business stays in business.
For some firms, the Amazon cloud or other public clouds may be all they need. For others, the risk of public cloud downtime is simply a risk they cannot afford.
Therefore, the million-dollar question is: What does your business require and where does your risk tolerance reside?
Get the latest news alerts: Follow WRAL Tech Wire at Twitter.