Amazon Web Service’s (AWS) Elastic Compute Cloud (EC2), based in Northern Virginia, went offline early this morning, taking with it a number of popular sites including, news aggregator Reddit and question and answer site (and TechCrunch darling), Quora.The Web hosting firm’s Relational Database Service (RDS), also based in Northern Virginia, and Elastic Beanstalk are down as well. according to reports on AWS’s Service Heath Dashboard.
Reddit.com was working in “emergency read only mode”
throughout Thursday, due to the problem with AWS, while geolocation site
Foursquare.com was back online Thursday afternoon, after being knocked
offline earlier in the day.
Amazon did not respond to Threatpost requests for comment.
Initial reports of problems began appearing just after 1:30 a.m. PDT (3:30 a.m. EST) on Thursday. The company said that it was investigating latency and error rates in their elastic block storage (EBS) that may have produce “connectivity issues” on the U.S. east coast. Just under an hour later, the cloud-based web hosting giant confirmed that they were indeed experiencing connectivity errors impacting their EC2 and RDS databases.
Around 3 a.m. PDT (6a.m. EST) Amazon confirmed increased error rates impacting their Elastic Beanstalk APIs and consoles.
For users of the many Web-based applications that rely on Amazon’s AWS infrastructure, the result was uncertainty. In the hours following the initial reports, Amazon’s service health dashboard pushed out a few optimistic updates an hour, reporting latency recoveries and varied degrees of progress, but no reports full service restoration. However, the EC2 and RBS “current status,” icons switched from a mild, yellow, indicating “performance issues” to the more ominous, red, which indicates “service disruptions” at around 11 a.m. EST. An update issued around that time was equally pessimistic, admitting that, “Despite the continued effort from the team to resolve the issue we have not made any meaningful progress for the affected database instances since the last update.”
AWS reported Thursday afternoon that progress was being made, but the issue remains unresolved. The company would not estimate when the problem will be resolved.
The wide ranging outage comes as something of an oddity as AWS is
generally seen as a stable web hosting service. In fact, in November of
last year, while those for and against Wikileaks were launching a slew DDoS attacks at any and all services deciding to host, unhost, support, or not support wikileaks, it was to Amazon’s EC2 web hosting service that the now infamous whistle-blowing site fled, because EC2 was deemed big enough and stable enough to withstand the attacks.