Most of the websites and services affected by late last week’s Amazon Web Services outage are back up after a little IT wizardry and a lot of consternation. But even as I write this, the North Virginia AWS location that caused all this trouble still isn’t back at 100 percent functionality. The Internet has largely put itself back together. But I’m afraid the widespread and very public failure of Amazon EC2 may have caused irreparable damage to cloud computing’s reputation going forward.
I don’t seem to be alone in this notion: Computerworld and Informationweek both express an urgent need to get Amazon Web Services up and running again. But more than that, this outage has caused a crisis of confidence. Never before in the short-but-eventful history of the web has there been a single outage that took out so many popular, highly used websites all in one go.
The fact that the problems were limited to the North Virginia availability zone mitigated the effects, but it’s still causing many doubts as to the security and reliability of Amazon’s cloud infrastructure. Given that AWS is often looked up to as the market and innovation leader in the IaaS marketplace, many are taking the failure as a sign of weakness in the very concept of the cloud.
Obviously, that’s a fairly extreme reaction. But I know for a fact that personal favorite web hangout Reddit is looking to take its million-plus monthly visitors and go to a competing hosting platform. And I doubt it is the only one.
With the Sony PlayStation Network still dark after a hacker assault, the cloud is not having the best month PR-wise. But make sure to continue watching TalkinCloud for further developments.
Follow Talkin’ Cloud via RSS, Facebook and Twitter. Sign up for Talkin’ Cloud’s Weekly Newsletter, Webcasts and Resource Center. Read our editorial disclosures here.
The 
I simply have to remark “are you kidding me?” I guess we all knew this defining moment in the reliability of cloud computing was coming – just didn’t expect it in lat April 2011. Otherwise I’m speechless…just sayin…
harrybbbb
Harrybbbb speechless? Say it ain’t so. Side note: I’ve been on the phone with our own web integrators to discuss backup/restore strategies in case our cloud provider ever went dark… Scary moments in recent days for some selected Amazon Web Services customers.
-jp
Overall, cloud computing offers too many benefits for this outage to have a lasting effect on adoption. Any one of the AWS customers could be in a dedicated environment and have the same outage. In fact, most would argue that anyone of them would be more likely to have a big outage if they were operating in a dedicate infrastructure. The difference is when a big cloud provider has an outage it affects several customers and those customers feel little control over the resolution or visibility into the resolution process.
A similar situation exists in the airline industry. Flying is far safer than driving your self, but the loss of control is unnerving to many passengers. It reminds me of when the airplane lands and the captain tells the passengers that the safest part of their trip just ended.
Mitchell: Your airline industry analogy aligns with my thinking. When Intuit’s cloud went dark in mid-2010, I compared a SaaS outage to airplane crashes vs. a PC outage to car crashes. Plane crashes (cloud outages) are far more terrifying but far rarer than car crashes (PC outages). But here are deeper reasons why:
http://www.mspmentor.net/2010/06/17/intuit-saas-outage-comparing-car-crashes-to-plane-crashes/
-jp