I've been mulling over the outage of Sony's PlayStation Network (PSN) for the past several days, searching for observations to put it into perspective alongside Amazon's recent cloud outage. Now I have my answer: There is no comparison.
PSN being offline since April 20 is a much bigger deal. "Cloudpocalypse continues into sixth day" is how the well-turned headline on the Website RCR Unplugged put it.
So why has L'Affaire Amazon apparently been deemed more serious, at least judging by the online chatter thus far? Both are black eyes for the cloud, if I can mix a metaphor. I think it has to do with differences in the services' user bases.
The Amazon EC2 hordes seem to have been quicker off the mark to scream bloody murder. While I can well imagine it must be hell not to be able to notify Foursquare that one has become "mayor" of kiss my something or other, I must also admit that Amazon's customers are indeed a large and diverse group of profit-making businesses. Their customers were unhappy, and they let Amazon know.
On the other hand, PSN's crash can be viewed as a point failure, in that it affects a single user base, albeit a base of 70 million users.
Seventy million! Who knew there was a cloud network in operation that, prior to fizzling last week around the same time Amazon's did, had reliably hosted millions of multi-user sessions of "Call of Duty: Black Ops" and "Killzone 3"?
Isn't this proof of cloud computing's robustness? As the parent of a 12-year-old boy who first clued me into the outage, I've seen quite a few of these sessions, and they're pretty impressive, both from the perspective of reliability (always up, until, of course, recently) and lack of latency (those guns shoot fast).
Sure, it all went to virtual hell sometime between April 17 and 19, when a hacker stole user names, passwords, and possibly Mommy's credit card data, too. ("While there is no evidence at this time that credit card data was taken, we cannot rule out the possibility," Sony stated in a blog post.)
More troubling for PSN's immense user base -- which, according to common sense, must include quite a few adults -- is that Sony doesn't seem to have a clue when it'll be fixed.
This stands in contrast to Amazon, which, as I recall, was reasonably forthcoming on its issues. (Check out status reports from the outages on Amazon's AWS Service Health Dashboard.) I should also note that, while the Amazon cloud went down April 21 and it took five days to restore full and complete operations, much of the service was pretty much back up, after a couple of days.
My broader point, though, is that Amazon has clearly designed its cloud to be robust and self-correcting. True, an unforeseen cascading storage failure seems to be at fault -- but that's not because Amazon didn't attempt to cover its bases.
The way Sony is flailing around, one has little choice but to assume that, while the company undoubtedly architected its cloud to run well -- I mentioned its great response time above -- it probably didn't plan on hacker attacks.
Hey, guys, haven't you heard that BC/DR -- business continuity and disaster recovery -- is the big new IT buzzphrase?
I give Sony the benefit of the doubt when it comes to its demurral on when PSN will be back up; it doesn't want to restart until it's sure the network won't go down again.
I'll close with some words from the PSN user I know best. A few days ago, I was pressing my son about why he seemed so blasť about the outage. "I thought it was gonna be back up soon," he told me Tuesday evening. "I didn't think it was a big deal. Now it is."
ó Alexander Wolfe is a technology editor focused on microprocessor and high-end software technology, as well as networking, storage, and cloud computing. He can be reached at firstname.lastname@example.org.