GMail Fail, Google's Future, and the Cloud
by Tony Kontzer
Last week's big Gmail outage was not the end-of-the-world scenario cloud computing detractors would have you believe, but it provides an ideal opportunity to consider two important implications.
The most obvious of these is what it tells us about Google. Let's face it, Google is the Web's answer to Teflon: No matter what problems it has, or mistakes it makes, it retains its standing as the most important and beloved site on the Internet. Not only does it continue to be the de facto online research tool for everyone from elementary school students to job-seekers to NASA engineers--no one I've talked to has switched to Microsoft's Bing--it's fast becoming the most beloved provider of Web-based productivity and collaboration tools, of which Gmail is obviously a critical one.
Despite this universal love and admiration, however, when Gmail is brought down by router failures, as Google says was the case, it brings into question whether the company can continue to support the ever-increasing load it's placing on its IT infrastructure. Sure, it's always excelled in delivering search results simultaneously to millions. Of course, it's delighted us all with its fast-performing mapping and satellite imagery toys. And yes, it's been getting a lot of love for offering up rock-solid tools, via Google Apps, that consumers increasingly want to access from multiple Internet-connected devices. But as it has continues to expand its footprint, to its Chrome browser, Android OS and beyond, it's fair to start wondering if it will ultimately start to buckle under the weight of its ambitions.
Business history is littered with the remains of companies that diversified to the point of detriment, and at times I worry that Google may be taking on more than it can handle. I do have faith that the plethora of brilliant engineers at Google has a handle on this issue, but it certainly warrants watching. (Adding insult to injury, Google's outage made Twitter look less bad, too.)
The potentially larger implication of the Gmail failure (the third such outage this year) for IT shops is what it may mean for cloud computing. I'm not suggesting that because Gmail goes down for a couple of hours, CIOs should start slashing planned cloud investments from their budgets. What I am suggesting is that any CIOs who brush off such incidents as consumer concerns that don't affect IT should seriously consider rethinking their positions.
Admittedly, there are still very few IT organizations of any size that rely on Google for any IT resources. But that could easily change as Google continues its strategic path to offering technologies that are enterprise-ready. And even if Google never succeeds in that pursuit, its experiences in the cloud provide a public test-bed for the readiness of the cloud. If a company with a data center as large and powerful and a user base as vast and demanding as Google's can't prevent such failures, how on earth is the typical IT shop going to ensure that its mobile sales staff won't lose access to customer data at critical moments?
Here's what I think CIOs should take from all of this: Running an effective cloud is a complicated undertaking that can stump even the most advanced IT shops. Does that mean all this talk of simplified IT and 24-by-seven access to data is a bunch of hooey? Of course not. But before you leap into the cloud, you better be damned sure of what you're getting into, because as sure as the sun setting in the West, if your cloud isn't reliable, you'll have a user revolution on your hands.
And no CIO wants that.