The outage on Amazon's European sites yesterday was due to a hardware failure, not hackers, the company has said.
The UK, Italian, German and Spanish sites went down for about half an hour on Sunday night.
According to internet monitoring firm Netcraft, Amazon's service health dashboard reported "elevated error rates and latencies for the EC2 APIs in the EU-WEST-1 region". Netcraft said many other sites may have been affected such as Amazon accounts for more than a third of all web-facing computers in Ireland.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
The Anonymous group behind Operation Payback had intended to carry out another attack against Amazon after last week's attempt failed to have any impact on Amazon.com. However, the websites and IRC servers operated by Anonymous suggest that the current target is still mastercard.com.
"The brief interruption to our European retail sites earlier today was due to hardware failure in our European datacentre network and not the result of a DDOS attempt," a spokeswoman for Amazon told Reuters.
There has been speculation that the attacks were in retaliation for Amazon's decision to remove Wikileaks from its servers.
Visa and Mastercard were the most recent sites to be targeted by hackers following the firms' refusal to process Wikileaks payments. PayPal and the Swiss bank PostFinance have also experienced problems.
A Twitter post from AnonOps, a group describing itself as fighters for internet freedom said, "WE REPEAT: We can not attack Amazon.com. To some journalists: check your sources well please."
Real-time performance graphs for these Amazon sites can be viewed here >>
Computer Weekly says...
Irrespective of whether the outage at Amazon was a DDOS or a hardware failure, the downtime is a major embarrassment and shows the vulnerability of public cloud infrastructure. Either way, Amazon web services has failed its customers, and its failings have demonstrated that even a public cloud service can have a single point of failure.