Vincent Gable’s Blog

October 14, 2009

Misunderestimating the Cloud

Filed under: Programming,Quotes,Security | , , , ,
― Vincent Gable on October 14, 2009

Recently, a Microsoft datacenter lost thousands of mobilephone user’s personal data.

A common response to this story is that this kind of danger is inherent in “cloud” computing services, where you rely on some service provider to take care of your data. But this misses the point, I think. Preserving data is difficult, and individual users tend to do a mediocre job of it. Admit it: You have lost your own data at some point. I know I have lost some of mine. A big, professionally run data center is much less likely to lose your data than you are.

Ed Felton

It’s easy to convince yourself of this anecdotally. Look around you, how many people people that you loosely know on Facebook have you seen complain about losing all their contacts when they lost their phone? I’ve seen at least a dozen such announcements. But nobody I actually know has been affected by this recent fiasco, or complained about losing contacts in any other “cloud” failure.

But people have a bias to overestimate risks they can’t control, and underestimate risks they can control. So we reinvent the wheel, and lose our own data ourselves.

Hey, I do it too. Actuarially, I really should be paying wordpress.com to manage this blog.

2 Comments »

  1. Failures of mass destruction:

    My disk dies, my backup disks all die, and I lose.

    Their disks die, and all their backups die, and LOTS of people lose.

    What is the cost of one failure in the two scenarios?

    PS. No-one I know voted for Bush. Doesn’t mean it didn’t happen.

    ED: fixed tense of die at commenter’s request

    Comment by Perry Wagle — October 14, 2009 @ 9:32 pm

  2. Perry, you are right to be skeptical of anecdotal evidence. I wish I had hard numbers on cloud failures, but I don’t at the moment. Everything I’ve read about personally-managed backups is pretty grim though,

    44 percent of U.S. adult computer users overall indicated that they
    have lost important data or digital files stored on their computer or
    laptop…

    My understanding is that “broken clouds” are very costly, but also much less common than broken PCs. So in general, everyone is safer with a very few massive data losses than with lots of individual-scale data losses. Of course, without hard data, that’s just “he said, she said”.

    What is the cost of one failure in the two scenarios?

    From my perspective, I think they are pretty much equal. I host my blog, and I screw up, result: all my writing is gone. I let wordpress.com host my blog and they screw up, result: all my writing is gone. Yes, other people’s stuff too, but I don’t see how that hurts me.

    From a societal perspective, I just don’t think the cost of one failure is a useful metric — we need to consider the chance of a failure, and weigh the expected cost. (Otherwise it’s like concluding motorcycles are safer than airliners, because the death-toll of a single crash is 100x lower.)

    Comment by Vincent Gable — October 14, 2009 @ 10:22 pm

RSS feed for comments on this post.

Leave a comment

Powered by WordPress