News

Amazon Cloud Outage Cripples Multiple Social Media Sites

• By Jeffrey Schwartz
• April 21, 2011

Amazon Web Services (AWS) experienced outages on Thursday that Amazon claims are due to connectivity and latency issues in its Elastic Compute Cloud (EC2) and Relational Database Service.

According to Amazon's Service Health Dashboard, the outages began early Thursday morning at the company's Northern Virginia datacenter and continued throughout the business day.

Affected AWS customers include Reddit, HootSuite, Foursquare and Quora, according to published reports. As of 6:00 p.m. Eastern Daylight time Thursday, only Foursquare's service appeared to be restored.

Of those four sites, only Reddit attributed its service disruption to Amazon's outage. According to a notice on Reddit's site: "Reddit is in 'emergency read-only mode' right now because Amazon is experiencing a degradation. They are working on it but we are still waiting for them to get to our volumes. You won't be able to log in. We're sorry and will fix the site as soon as we can."

Amazon pointed to a single availability zone in the Virginia datacenter as the problem. According to a status message at 4:48 Eastern time:

"All other Availability Zones are operating normally. Customers with snapshots of their affected volumes can re-launch their volumes and instances in another zone. We recommend customers do not target a specific Availability Zone when launching instances. We have updated our service to avoid placing any instances in the impaired zone for untargeted requests."

Service at Amazon's Northern California datacenter and its facilities in Europe and the Asia Pacific appeared to be running fine, according to the dashboard. Experts say customers can avoid the impact of such problems by selecting multiple availability zones or, even better, multiple cloud providers.

"If your business relies on a Web site to be up, why do you allow a failure in a single availability zone to shut down your business?" wrote Scott Sanchez, security and privacy officer at ScaleUp Cloud, in a blog post. "There are so many tools out there at this point to simplify deployment, scaling and resiliency across multiple availability zones or even across multiple cloud providers -- frankly, you have no excuse."