The more we rely on technology, the more devastating it is when there's a network or systems failure. And no matter how many redundancies and failovers service providers and private businesses build into their backup plans, it's amazing how one small blip in the wire can still bring everything to a screeching halt.
Sometimes it's just an annoyance. Imagine my husband's extreme irritation when, three-quarters of the way through the season finale of Sons of Anarchy, our television screen suddenly displayed "Attention: Your Explorer settop is NOT AUTHORIZED for use. Please call: (800) 266-22XX."
More serious outages cause business disruptions, which can result in reduced productivity, damage to reputation, breaches in security, and lost revenue. According to network services company MegaPath, the average cost of downtime equates to $212,200 per hour -- a figure that can skyrocket for web- and network-based businesses. In August, the Amazon.com website went down for a period of 30 to 40 minutes, costing the company between $3 and $4 million.
In worst-case scenarios, crippled networks can render emergency services and public safety communications systems useless. The countrywide outage of Canada's Rogers Wireless in October left millions of customers without access to 911, with city officials advising people to use neighbors' landlines or phones at nearby stores in case of an emergency.
No company is immune, no matter how sophisticated its technology. In 2013, some of the largest and most well-known technology companies in the world experienced downtime because of network failures. Companies such as Google, NASA, Facebook, Netflix, Twitter, and others all reported outages stemming from some problem in the network.
Whether downtime is caused by a faulty piece of hardware, a software bug, a configuration error, or a denial-of-service attack, it causes the same response in customers, employees, users, and the IT professionals responsible for that network: Sheer panic.
So don't sit idly by and wait for the next time a problem brings your network to its knees. Take a gander at the top network outages of 2013. Learn what you can from these companies' misfortunes and try to avoid the same fate.
Photo by Marcelo Graciolli.
Connectivity issues haunt AWS
Amazon Web Services prefers to experience major downtime on days of significance -- in 2012, it was Christmas Eve, and this year it was Friday the 13th. This time, "network connectivity issues" in a single availability zone within Amazon's now-infamous US East region were the culprit, Amazon said, bringing down some Amazon cloud services as well as sites like Heroku, parts of Github, and CMSWire. Although repeated failures in Amazon's US East region seem to be related to aging infrastructure, David Linthicum of GigaOm explained, "The network issue that caused the problems on Friday could easily be worked around with the right approaches and automation capabilities."
Food stamp service knocked out
There's nothing like that sinking feeling when your payment card is declined. That's what residents of 17 states served by Xerox Corp.'s electronic benefits system were feeling when a routine test of backup systems caused a failure lasting most of a Saturday -- prime grocery shopping time. Recipients of programs such as SNAP and TANF in states including California, Massachusetts, New Jersey, Ohio, and Texas were able to access their benefits only through an emergency voucher process until the entire system was rebooted.
Boo hoo for Yahoo email
Earlier this month, a Yahoo Mail outage cut the lines of communication of users for an entire work week, even longer for some. Afterward, Yahoo CEO Melissa Mayer issued a public apology on her Tumblr account attributing the problems to a hardware failure and noting that the failure affected about 1% -- or approximately 1 million -- of the service's users. Yahoo was criticized for its lack of communication with users during the episode, which came on the heels of a redesign that met with mixed reviews.
OMG, Google goes dark
A glitch somewhere inside Google caused its entire network, including search, YouTube, Gmail, and Google Drive to go dark for a short period in August. Although the outage lasted only between two and five minutes, analytics firm GoSquared estimated that worldwide traffic dropped 40% during that time. Google has been experiencing reliability issues with Gmail in particular, but this was the first time its entire system was affected, and the repercussions were enormous.
Nasdaq hits the brakes
The Nasdaq stock exchange was forced to halt trading for three hours in August, leaving investors and brokers cooling their heels. The shutdown was prompted by a disruption in the connection between the NYSE Arca network and the data processing subsidiary of Nasdaq. When the Nasdaq backup system kicked in, a software bug prevented it from functioning correctly, resulting in a total failure.
Photo by bfishadow.
Faulty optical switch boots NYC offline
Technical issues with a Level 3 Communications fiber optic switch caused a major Internet outage to Time Warner and Cablevision customers in the northeast this fall. While Cablevision was able to recover its service by rerouting traffic fairly quickly, Time Warner customers in an area including New York City, as well as many corporate customers, had to wait for Level 3 to fix the offending switch, which took more than 24 hours. The outage fortunately occurred on a Saturday, avoiding a great impact on business customers. However, Time Warner had just recently signed a deal to provide Internet access to the more than 7,000 Starbucks locations in the US.
Communications silenced in Canada
Rogers Communications, Canada's largest wireless phone company, lost service to a large number of its customers nationwide for several hours in October. It was the company's second outage in 24 hours, after the Rogers.com website failed to function earlier. The wireless failure proved far more serious, however, leaving customers unable to make emergency calls from mobile devices. Rogers advised those in need to call 911 from landlines, and the City of Calgary suggested that citizens without landlines borrow phones of neighbors or local businesses if necessary. The outage was ultimately attributed to an unspecified software problem.
Network Solutions customers get double-whammy
DNS registrar and hosting service Network Solutions lost service for the third time in the year in October, angering some customers enough to switch providers. The company reported it was experiencing problems sending email because its servers had been blacklisted and messages were being marked as spam. In the same day, an undisclosed DNS issue caused disruptions for several business customers, who vented their frustration via social media platforms.