In May, British Airways became the latest household name to
defend itself over a major outage that led to chaos for thousands of travellers
over the May bank holiday weekend.
Datacenters – particularly those operating for businesses
with large numbers of customers – are inherently risky. Most are vulnerable to
human error, software or hardware quirks, theft and cyber attacks, or even
natural disasters. Let’s take a look at three prominent examples, starting with
BA.
1.
British
Airways
BA originally attributed the global failures of its ticketing, baggae
handling and operational systems to a power surge. In actual fact, it was the
mistaken
actions of a single datacenter employee that led to the cancellation of
around 800 flights from London Gatwick and Heathrow airports on one of the
busiest weekends of the year. An internal investigation revealed
that no IT failure occurred and that all software ran without issue. An act of
human error is estimated to cost BA approximately €61m (CHF 66.93m) in
compensation.
BA might be exceptional in the
scale of the effects of their downtime, but they are not alone in experiencing
outage problems. In 2016, Delta was forced to cancel thousands of flights, and,
more recently, Lufthansa and Air France were briefly unable to board passengers
due to global systems failures.
2.
Vodafone
The theft of switch equipment from a
Vodafone datacenter in Basingstoke, UK left hundreds of thousands of customers
without calls, texts or mobile internet services in February 2011.
Theft is a common threat to datacenter
operations. Other notable examples include the ‘Ocean’s 11’ datacenter robbery
in London, where thieves disguised as police officers stole over $4m (CHF 3.84m)
in Verizon equipment in 2007. And, in 2008, the Financial Times’ website was
taken down by the theft of datacenter components at a Cable & Wireless site
in Watford, UK.
3.
Datagram
Buzzfeed, the Huffington Post, Gawker and
Gizmodo fell victim to Hurricane Sandy in 2012 when the basement of AT&T’s
global network operations center (GNOC) in New Jersey, USA was flooded during
the storm. Proving that no location is completely safe from natural disasters,
the location was chosen in part due to the low likelihood of hurricanes and
earthquakes.
Storms have also been the source of outages
for HBOS, a UK banking group, which left customers unable to use ATMs or access
their accounts online for six hours in November 2009.
Datacenter managers must be simultaneously proactive –
prepared for perceived risks – and reactive – ready to respond to the unknown.
According to their cloud product manager, Google even
coat
their trans-Pacific underwater cables in Kevlar to protect against shark
attacks.
It’s impossible to plan for all the risks, as all the
examples above have shown. You can, however, take steps to make sure your staff
are kept up to date on their product knowledge. That’s where simulation comes
in.
Previously reserved for surgeons, pilots and astronauts,
simulation technology is becoming increasingly available to other industries.
Simulcation Datacenter is the result of
a ground-breaking partnership with IBM, empowering IT staff to experiment with
their decisions and test-drive leading architecture in an engaging environment.
Preregister now to be one of
the first to receive an account.
About SimulcationWe believe learning
shouldn’t be laborious. Simulcation Datacenter is an engaging, competitive
environment that simulates the complex challenges faced on a daily basis by IT
infrastructure professionals. Preregister now to be one of the first to
experience it – or contact us to find out more.
