It’s often said – partly with tongue in cheek – that if Maslow’s famous hierarchy of human needs were revised today, internet connectivity would feature at its foundation. Look around any office space when the internet goes down, or a busy train carriage that moves out of signal range. The palpable sense of aimlessness highlights just how reliant we have become on connectivity.
Yet, as pervasive as the internet may be, there are still rare instances where it can be lost for more than a brief moment. Only in December, a power outage at a Manchester-based data centre brought hundreds of businesses to a standstill on one of the busiest traffic days of the year. The knock on commercial effects of such occurrences can be catastrophic for both providers and business customers.
The typical cause of a mass outage varies. According to the Ponemon Institute, the three most common causes are power supply failure (25%), cybercrime (22%) and accidental human error (22%). In this post, we look at some of the most notable network outages in recent years, what caused them and how to be prepared in the event of a future reoccurrence.
1. Dyn cyberattack (2016)
Countries affected: United States and Europe
Impact: Millions of users of popular sites including Airbnb, Amazon, BBC, CNN, eBay, Netflix and Twitter
What happened: It’s often easy to forget just how interconnected the web is, and how reliant the online information and services we use daily are on shared underlying protocols. This was clearly exemplified in October 2016, when multiple distributed denial-of-service attacks (DDoS attacks) targeted systems operated by Managed Domain Name Systems (DNS) provider Dyn.
Thought to be the largest DDoS attack in history, the outbreak originated from malware vulnerabilities in a large number of internet-connected devices such as printers, IP cameras and baby monitors.
As a result, hundreds of popular websites were taken offline for hours, bringing chaos to users and costing businesses millions in lost revenue.
2. British Airways IT failure (2017)
Countries affected: United Kingdom/global
Impact: Thousands of passengers globally
What happened: Everyone at some stage has experienced the frustration of an airport delay. Perhaps an adverse weather event or traffic signalling problems can result in mild confusion and lost hours at the airport terminal. For British Airways customers in May 2017, the situation was far more dire.
A global IT crash resulted in the airline’s entire fleet being grounded, leaving thousands of passengers stranded over one of the busiest weekends for travel in the UK. So, what caused such widespread chaos? An IT maintenance contractor accidentally switched off an ‘uninterruptable power supply’ at a key data centre, damaging the server during the subsequent power surge.
Along with the mandatory compensation payments and the hit to revenues and profits, the obvious consequences for the airline’s reputation have had a lasting impact.
3. Amazon Web Services (2017)
Countries affected: Global
Impact: Millions of cloud service and website users
What happened: It seems almost unfair to single AWS out here, as nearly every cloud platform has suffered some form of outage in recent years. Yet, it’s only when one of the biggest goes down that you realise just how dependent we are on these cloud behemoths.
In this particular instance, Amazon’s S3 web-based storage service experienced high error rates, leading to either partial or full outages on a number of popular websites, apps and devices. Sites affected included Quora, Business Insider and Slack, among others. Originating in its US-East-1 data centre, it’s thought the issue could have been avoided had developers spread their applications over different facilities.
Indeed, this wasn’t an isolated incident. The provider suffered similar outages in September 2015 due to software problems at a North Virginia data centre.
4. Vodafone data centre (2011)
Countries affected: UK
Impact: Thousands of Vodafone customers
What happened: Businesses invest millions each year in ensuring their data servers are safe from the multitude of online threats. Given the ever-evolving nature of cyber attacks, this is money well spent. Some businesses may, however, overlook the physical security aspect, which is equally important.
Mobile phone operator Vodafone learned this lesson the hard way after thieves broke into its Hampshire-based data centre using sledgehammers. The breach caused major service interruptions, leading to a flood of angry complaints from customers.
5. A whole country without connection (2011)
Countries affected: Republic of Armenia
Impact: 3 million people
What happened: This is perhaps the most humorous of our examples – unless, of course, you’re Armenian. It’s not often that an entire nation loses connectivity, yet that’s exactly what happened to the small former Russian republic in 2011 when a 75-year-old woman digging for scrap metal damaged land cables and thereby severed most connectivity across the nation. To rub salt in the wound, the culprit wasn’t even Armenian – the ‘spade-hacker’ was conducting her excavation in neighbouring Georgia!
What can you do to be prepared?
The explosive growth of mobile usage and Wi-Fi connectivity over the past decade has brought a new set of connectivity needs and challenges for businesses and network providers alike. While a great deal of work has been done to secure network infrastructure and develop software that can guard against emerging threats, the chances of similar outages happening in the future still exist.
While it must be emphasised that such instances are rare, the sudden and unpredictable nature of outages can highlight even the slightest network or security vulnerabilities. It’s therefore essential that businesses and providers do their due diligence in preparing for them. Risk assessment, planning and provisioning are a given for most organisations, but to be truly prepared businesses must ensure their IT architecture has flexibility, security and resilience built in.
Colocation data centres can provide a simple and effective route to greater network resilience. By colocating, a business can have multiple points of presence across different facilities, meaning less reliance on any single data centre being fully operational. Adding world-class digital and physical security provides an extra layer of defence against emerging threats, whether malicious or otherwise.
No business can have the luxury of knowing the challenges tomorrow will bring, but with the right IT systems in place they have the greatest chance of keeping their business – and those dependent on it – online. They’d be lost without it.
Michael Rabinowitz is director of marketing & strategy, connectivity at Interxion. He has over 20 years of experience in the telecommunications market with experience of infrastructure and service development with telcos across the globe.