Tuesday, September 8, 2009

Cloud Attraction Survey

I labeled this blog The Cloud Option because of my belief that the best reason to build an application with a cloud architecture is to manage application demand risk. The cloud option allows you to align application demand with infrastructure supply to protect against a demand forecast that is certain to be inaccurate. While I believe my value hypothesis will prove to be correct in the long term, I think there may be far more basic attractions associated with the near term value driving cloud demand.

With that in mind, I have built a short (12 question) survey that I offer to AWS users (as I believe AWS represents the most successful, high demand implementation of cloud computing by far thus far). If you are a current AWS consumer, please take 2 minutes to fill out the survey. I'll post the results on the blog in a week or two. Thanks for helping me assign value to the cloud option!

Wednesday, September 2, 2009

Latest Gmail Outage Again Fuels Cloud Computing Luddites

By Steve Bobrowski

What's a Luddite? I remember looking this up once, and here's one of the definitions I found at Webster's:

Change scares people, making them feel uncomfortable and uneasy for many reasons. But in the world of technology, we must embrace change, for it is inevitable and fast-paced. And IT change usually happens for the better, not the worst.

In this context, my news wires on cloud computing today have been flooded with countless stories written by a bunch of, well, Luddites. Those insisting that the latest Gmail outage is proof that cloud computing system outages threaten the cloud computing paradigm shift. Really?

The truth is that system outages are a fact of life. We all hear about the public system outages, but we rarely hear about those that occur behind the firewall. Rather than trouncing cloud services when they go down, shouldn't the focus be on how long it took them to return to service and then comparing the impact of this event to similar on-premise outages?

For example, when was the last time that your organization's email system went down? Did your IT department have the training, staff, and resources to quickly identify the problem and then fix it? My personal recollection of an internal email system outage: lots of squabbles and finger-pointing among the parties involved, all leading to two days without email and some lost messages. Never mind the ripple effects of lost time and money while our IT staff needed to suspend work on other internal projects.

Google's team of highly-specialized administrators took less then two hours to fix things and I didn't lose any of my mail! In my experience, that's outstanding service. Furthermore, fixing the outage did not require work from any of my company's in-house resources, which were free to continue being productive on internal projects that lead to revenue generation.

Should your organization worry about relying on a cloud application or cloud platform? The answer is simple—applications should reside where it makes the most sense. In some cases, cloud wins, in others, data center wins. But the trend is undeniable that more and more enterprises are outsourcing common business applications such as email and CRM to the cloud because it provides their workers with a better service and more time to work on core business functions.

In summary, don't let the Luddites scare you—the inevitable world of utility-based computing will improve the enterprise's standard of living in many cases, not the opposite, as they would have you believe.