Web 2.0 companies -- collaborative software, social networks, and the like -- love to outsource their computing infrastructure. That way they can scale operations as necessary without buying the hardware and software that would suck up money faster than a banker on Bailout Day.
And it's great when it works. But as the Amazon (AMZN) cloud outage shows, when it breaks down, the experience is brutal and leaves businesses out of control, waiting for someone else to fix the problem.
Amazon had a significant failure in its Virginia facilities. The company first publicly noted the problem at 1:41AM Pacific today. By 11:09AM, it still couldn't provide "an estimate that we believe is close to accurate," but said that "all-hands are on deck to recover as quickly as possible."
This cloud? No silver lining
When you're a cloud vendor, your problems are your clients' problems. The cloud outage brought a number of web sites down, including such high-profile services as Foursquare, Reddit, and Quora. A day's outage could translate into a 0.3 percent reduction in annual revenue (1 day out of 365), and it's difficult to say what the impact on customer relations going forward might be.
Cloud problems go well beyond the occasional disruption. Iron Mountain recently said it would shut down its public cloud-storage business, becoming the third company to do so. At least it gave customers a year advance warning -- unlike Mozy (EMC), which last year announced an abrupt and immediate end to its unlimited storage plan and a simultaneous increase in prices.
It's a basic problem with outsourced infrastructure. Your company becomes hostage to the vendor. Whenever there's a technical incident or a sudden change in strategic direction, your company can't easily shift service providers. That takes complex migration planning.
In a phone conversation, Aaron Levie, CEO of Box.net had discussed his company's recent $48 million D round of funding. A big reason for the additionalinvestment was to further expand the company's data centers:
If you can imagine the kinds of customers we're dealing with -- regulated businesses like banks and pharmaceutical companies, Fortune 500s -- we need the right types of infrastructure and processes. Our customers are coming to us as a technology provider and vendor. We've invested significantly so we can support and directly be responsible for the success of our customers' technology.Box.net is a specialized type of cloud services provider. Then again, almost every Web 2.0-style business is. Ever since the beginning of technology outsourcing, the experts have said to push to vendors what aren't core functions. But if you are a cloud vendor, is it really wise to outsource your cloud infrastructure? Something that could stop business in its tracks by going down sounds like a core function to me.
Will cloud computing disappear? Absolutely not. But over time I'd bet that companies will look at the tradeoffs more carefully.
[Update: About those trade-offs -- Amazon's crash permanently destroyed some customer data. What company can afford to possibly lose data because a cloud provider doesn't use off-premises backup?]
- Apple Wants To Store Your History in the Cloud
- Livedrive's New Online Storage Secret: Paper Printouts
- The Epsilon Email Break-In: A Bad Break for The Cloud
- Amazon Opens Its Music Locker To Close Out Apple and Google
- Battered by Criticism of Its Board, HP Sticks Its Head in a Cloud