A rash of thunderstorms across the United States recently left widespread damage across much of Maryland and the Virginias. Power outages left many in the dark for days, including Amazon's servers at its main data center in Ashburn, North Virginia. This incident, while mainly effecting media availability, has some questioning whether the cloud is really the safest place to store information.
The federal government has been pushing for more cloud computing and BYOD (bring your own device) implementation for its workers to allow greater mobility and to cut costs. While the Obama administration has been an advocate of this plan for years, some are now worried that on top of IT professionals policing endpoint security, they now must divert extra attention to maintaining cloud data from outages or loss.
A rumble of concern
The U.S. government has been moving its key services and server facilities to the cloud for the last two years. During that time, email, web page maintenance and general data storage have all been incorporated into cloud solutions. The reasons for this include cutting costs and creating a centralized resource for all offices to use while improving communication between departments. Now officials are concerned what it's really achieved is threatening federal data protection with too much reliance on virtualized storage.
"Last week's powerful thunderstorms, along with the massive disruptions they caused, exposed some of the vulnerabilities of cloud computing," said U.S. Representative Mary Bono Mackof California, according to the Washington Post.
The powerful outage
The incident occurred when a series of severe thunderstorms rolled up the Eastern seaboard, producing damaging winds and hail with driving rain. These conditions led to many localized blackouts and extreme damage to residences and businesses, but it also knocked the Amazon Web Service (AWS) offline for a few hours.
The issue with AWS was that its Elastic Compute Cloud (EC2), which is hosted on the East Coast out of the Ashburn data center, lost its power sources and could no longer maintain a connection with other host servers, according to Fierce Enterprise Communications. The service, which hosts sites for industry leaders like NASDAQ and Netflix, went down for a few hours before power was restored and the server connection repaired.
While the federal government was not affected directly by this outage, as its data storage is handled by another vendor, it has caused key officials involved in the administrative processes of data protection to reconsider the current information strategy.
"The fact that Amazon, like any other data center-dependent business, is not bulletproof also raises questions about why its customers don't pursue a multi-cloud strategy," wrote Barb Dorrow of GigaOM in the Enterprise article. Dorrow stated that placing so much of a business' architecture in a single resource was asking for a business continuity problem, but it also wasn't fair for Amazon to blame its clients for lack of preparedness when it was AWS' fault in the first place that precautions weren't in place.
Another GigaOM industry specialist, Katherine Fehrenbacher, said that these power problems have led others to reconsider online storage as well, including vendors of the service. Apple, for instance, has announced that it will handle at least 60 percent of its data center power needs under its own volition so that, even in a severe outage, its iCloud and other services won't crash. Still, even if every cloud provider took itself off the grid, it wouldn't make an outage impossible. Moving toward a multi-layer solution is imperative for protecting data security.