An alleged lighting strike at Amazon's (NASDAQ: AMZN) Dublin data center triggered a chain of events that led to an explosion and fire that took out the power supply, and with it, the ability to switch seamlessly to backup power generators.

This version of events was disputed by Dublin-based energy supplier ESB Networks however. As reported by The Register, a spokesman at ESB told the publication that "the cause of this failure is still being investigated at this time but our initial assessment of lightning as the cause has now been ruled out."

ESB says that about 100 customers were affected, and all had electricity available either from the main supply or via standby generators in an hour.

What is clear though is that the electrical problems resulted in a total power outage and necessitated manually bringing backup generators online. Moreover, downed EC2 instances must be validated being before restored, taking about 12 hours just to restore 60 percent of the instances.

In all, it was at least six days before all services were back to normal. For businesses who have yet to learn from the Northern Virginia data center outage earlier this year, the inevitable conclusion is that cloud computing does not eliminate the need for contingency planning--either by failing over to a different availability zone within AWS, or to other cloud providers.

According to Amazon's Service Health Dashboard, all services within Europe are functioning normally at the time of writing.

For more:
- check out this article at ZDNet
- check out this article at The Register
- check out this article at CRN

Related Articles:
Cloud risks put in spotlight by Amazon outage
Questions raised over latest Amazon EC2 outage
Amazon apologizes for cloud outage, promises to do better