Skip to main content

"Human error" led to Amazon AWS outage

Amazon has come down with a case of the Sonys, admitting that its web hosting service was downed due to "human error", apologising for the outage and promising compensation.

A number of servers in the Amazon Web Services (AWS) cloud went down on April 21, some staying offline for two days. With many major sites such as Quora, Reddit and Foursquare using Amazon's cloud service, the damage and customer discontent was monumental.

Amazon said that the outage began when it tried to upgrade the network capacity of its datacentre in north Virginia, according to Datamation. The procedure for such an upgrade was not followed to the letter, resulting in Amazon staff accidentally rerouting traffic into the wrong area, which resulted in the servers overloading.

The problem was so bad that some AWS customers lost data. Amazon has not revealed how many were affected by data loss and whether or not these customers will receive additional compensation.

Amazon apologised for the downtime and for its poor communication to customers about the problem. It also promised ten days of free service for those affected by the outage. However, that may be a small token for those who lost business due to the outage.

This is the latest in a string of cockups by major technology firms. Sony's PlayStation Network was recently the victim of a huge security breach that resulted in customer details being stolen. Sony apologised for the downtime and offered 30 days free service to its Plus service as compensation.