Amazon Web Services experienced a service disruption that affected a single availability zone in its US East (Northern Virginia) Region (US-EAST-1) on Friday 13. The outage was reported on its
Service Health Dashboard at 732 AM PDT and resolved by approximately 1030 AM PDT. Key issues specific to the disruption include:
•Network connectivity such that affected instances were unreachable by means of a public IP address, but could communicate with other instances in the same availability zone using private IP addresses
•“Increased error rates and latencies for the EBS APIs”
•“Increased error rates for EBS-backed instance launches”
•Increased EC2 API latencies
•Impacts on EC2, the Relational Database Service, the Redshift data warehouse, load balancers and the simple email service.
Affected customers included Heroku and Github. The service disruption raises questions about the ability of Amazon Web Services to support enterprise customers, particularly since the Northern Virginia region has been site the site of several outages and technical issues over the last three years. Moreover, the negative publicity for Amazon Web Services come at precisely the time when the battle for IaaS market share is set to intensify given increased competition from the likes of Microsoft, HP, IBM, Google and Rackspace.