Lightning strikes Microsoft and Amazon data centres

Tue, Aug 9, 2011, 01:00

A €350 million Microsoft-owned data centre opened in Dublin two years ago because directors of the company liked the weather here.

But the centre became a victim of that weather at the weekend when a lightning strike brought some of its most cutting-edge cloud computing services crashing down.

The software giant opened its “mega-data centre” at Grange Castle in CityWest in September 2009. It is its first such facility outside the US and has been pivotal in Microsoft’s attempts to secure a bigger slice of the cloud computing market in Europe.

Cloud computing involves running software remotely in the company’s data centre – “in the clouds” – rather than on customers’ machines, and is set to heavily influence the future of both business and home computing.

One of the largest costs associated with operating such a service is cooling the enormous heat generated by the computer servers needed to store information belonging to potentially millions of customers.

At the time the facility opened two years ago, Microsoft said it had chosen Dublin as the location for the cloud centre in order to make use of ambient air to cool the facility “We love the Irish weather,” said Microsoft International president Jean-Philippe Courtois. “The cool climate of Dublin was a very important consideration for us.”

The company is probably a lot less enamoured with the weather here now. On Sunday, the lightning strike hit a transformer station in the CityWest facility and shut down the data hub that was being used both by Microsoft and online retailing giant

Amazon appears to have been the worst affected by the strike. In a statement the company said it had caused “an explosion and fire”. It stressed that while there was a backup generator in place, “the explosion was large enough” to have disabled some connections, rendering systems inoperable for several hours before being manually restarted.

Amazon’s cloud services were down for several hours on Sunday night and services across Europe were still facing problems yesterday because of the scale of the disruption. The company said it would take up to 48 hours to complete the restoration process.

Microsoft’s own service was also knocked out by the strike, but according to a Twitter update from the company early yesterday morning, the service was back within hours.

An ESB spokeswoman said the incident occurred at 6.15pm on Sunday and caused a power outage in the area for about an hour. However, she said power to Amazon was interrupted for “less than a second before an automatic supply restoration kicked in.”