It took cloud computing pioneer Amazon Web Services three days to resolve its outage last week fully, and a small proportion were unable to recover hosted systems altogether.
The outage was triggered by a "networking event" on Thursday morning, and affected customers whose systems are hosted at Amazon’s data centre in North Virginia. Both the Elastic Computer Cloud (EC2) server hosting and Elastic Beanstalk application hosting services were struck by increased latency and errors.
The company worked to recover lost instances for the following three days. By Sunday, it had recovered the "vast majority" of instances. However, the company said yesterday that "a small number of volumes (0.07% of the volumes in our US-East Region) will not be fully recoverable".
"We are digging deeply into the root causes of this event and will post a detailed post mortem," the company said.