Wow. No sooner did I finish writing about how the…
Wow. No sooner did I finish writing about how the Google and Microsoft outages were not a reason to lose confidence in the cloud, than Amazon went down. The online retail site—and its associated cloud services—were down for just under half an hour Monday afternoon. I stand by my assertion that the sky is not falling, but there’s more to using the cloud than just availability.
Amazon.com was the third major cloud service to suffer an outage in the last week.
Over on WindowsITPro.com, Paul Thurrott summed up the hysteria over cloud outages nicely. “And of course, the cloud computing doubters—who, like global warming doubters are increasingly at odds with reality—will argue that such outages prove that our move away from on-premises hardware and local storage is nothing but a temporary trend.”Let’s start with some perspective, breaking down the math like I did yesterday for Google and Microsoft. Amazon was down for about 25 minutes (although I’ve seen reports from 15 minutes to 40 minutes). In the grand scheme of things, Amazon was down for an infinitesimally small period of time. Depending on the estimate you go with, Amazon lost about $5 million in retail commerce during that timeframe—or about two percent of what it cost Amazon CEO Jeff Bezos to buy the Washington Post, or about two thousandths of a percent of his net worth.
Thurrott also pointed out the irony of how many users turn to Internet-based services like Facebook or Twitter to complain about cloud outages and declare the impending death of this cloud fad.
Availability is just one factor
The debate over cloud availability is silly. As I pointed out my post about the Google and Microsoft outages, local networks and servers are not impervious to outages, so the risk is essentially the same as it pertains to availability.