Traffic in Amazon (NASDAQ:AMZN) Web Services’ most heavily used data center complex, U.S. East-1 in Northern Virginia, experienced server issues due to an outage in one of its availability zones early last week. With Hurricane Sandy barreling toward the East Coast Monday, many websites might fear that similar disruptions might occur.
Last week, the outage effected the availability of several popular services, such as Netflix (NASDAQ:NFLX), Reddit, Pinterest, Foursquare and Imgur. A malfunctioning server seems to have caused the problem. 
We think that the reputation and position of Amazon as a leading cloud services provider is suffering from the recent outages, which have been caused by a multitude of reasons from inclement weather to malfunctioning servers.
The recent outage was similar to the one back in June 2012 which claimed Netflix and Instagram among its victims. The June outage was caused by an electrical storm and had its origin in the same data centers. There was another EC2 (Elastic Compute Cloud) outage in April 2011, which stemmed from EBS-related issues. This time too, the cause was “a latent bug in an operational data collection agent that runs on the EBS storage servers”.
EC2 and EBS, as part of the Amazon Web Services suite, are designed with substantial amounts of redundancy and fail-over capability. In addition to having local redundancy, the services are divided up into partitions called “Availability Zones,” and large customers can spread their EC2 instances out across multiple zones. Last weeks outage only affected EBS volumes in a single Availability Zone, so those customers running with adequate capacity in other Availability Zones in the US East Region were able to tolerate the event with limited impact to their applications.
We believe, that in spite of the lessons learned from last year’s outage, the automated procedures built into EC2 still do not protect against some kinds of failures. This could have a two-fold effect on the growth of the company’s web services business as existing customers could chose to instead have their own data centers or chose to side with another cloud service provider and new customers, deterred by the unreliability of the system may chose a competitor over Amazon.
Currently, the Cloud and Other Web Services contribute just 3% to our $222 estimate for Amazon. However, the company has been looking to diversify its revenue stream and cloud services is a major part of its growth plans. Given that a lot of Amazon’s customers are small businesses who cannot afford or don’t need to use multiple zones to back their services, the company could lose a major chunk of its customers in the event of another outage and with them, a potential revenue stream.
It competes directly with Google (NASDAQ:GOOG), Microsoft (NASDAQ:MSFT), IBM (NYSE:IBM) and HP (NYSE:HPQ) in the cloud computing market and will soon have another formidable competitor in Oracle(NASDAQ:ORCL).
- Summary of the October 22,2012 AWS Service Event in the US-East Region, Amazon Web Services, October 2012 [↩]