As Google Goes Down, Private Clouds Go UpAs Google Goes Down, Private Clouds Go Up
I've already made the point that there's no such thing as "private clouds." It's an oxymoron because cloud computing, by definition, happens when people plug into IT services offered from data centers that aren't their own and that are shared by others. That said, cloud architectures will indeed be popular in corporate data centers, even if we disagree on what you call them.
I've already made the point that there's no such thing as "private clouds." It's an oxymoron because cloud computing, by definition, happens when people plug into IT services offered from data centers that aren't their own and that are shared by others. That said, cloud architectures will indeed be popular in corporate data centers, even if we disagree on what you call them.There are a lot of reasons that CIOs are interested in private clouds (I'll use the term now that I've made my point). As I pointed out in "Private Clouds Take Shape," IT departments have long sought utility-like IT environments where computing resources and applications can be provisioned with greater efficiency. Other compelling reasons include security, regulations and governance, data control, vendor independence, application performance, and, as the past few weeks have demonstrated, system reliability.
There's been an alarming number of outages among cloud service providers recently. On July 20, Amazon's S3 went on the blink for eight hours, to the consternation of many customers. Since then, my colleague David Berlind has documented outages with Google Apps and Gmail, with Citrix's GoToMeeting and GoToWebinar, and then again with Google Apps and Gmail. David's headlines tell the story: The Week The Cloud Went Down; The Month The Cloud Went Down; and Oops, They Did It Again.
We know from experience that it's very hard to build IT infrastructures that scale to hundreds of thousands and millions of users with five nines availability, and we're seeing that play out in the public cloud. For years, IT departments have been willing to pay a premium for high availability systems because the business units they support can't afford the downtime associated with system glitches.
Knowing those two fundamental truths -- public clouds fail, but businesses demand systems that work -- we can make the following predictions with a high degree of confidence. First, cloud computing providers will scurry to improve the availability of their services. Second, new services and products will spring up that promise improved cloud availability and redundancy and backup when things go wrong. And third, businesses will build their own cloud environments where they're able to control the many variables that affect service availability. What private clouds lack in scale and hands-free management, they make up for in security and reliability.
Read more about:
2008About the Author
You May Also Like