AWS Outage Shuts Down AP Classroom

On October 20, Amazon Web Services experienced a massive outage, disrupting major companies globally. As a result, over a hundred commonly used services from different sectors were affected. On top of Amazon services like Alexa and Prime Video, there were also platforms that many students rely on or use often, such as Canvas, AP Classroom, and Roblox.

In particular, AP Classroom was completely inaccessible during the outage because it uses AWS. The rigorous BIFU curriculum necessitates regular use of this website for assignments, tests, and other academic purposes, so not being able to use it greatly hampered the productivity of students.

“I was pretty happy because for the moment it saved me from AP classroom, but at the same time it just meant that I would have more work later,” says Aaron Ely (10).

AWS is an example of cloud computing, which is providing computing resources such as data storage and servers over the internet. Companies can get the resources they need without the high costs of owning and maintaining their own physical computers.

In 2006, AWS was launched as a platform that provided this service on-demand, only charging for what was used. This model, when compounded with other benefits like global infrastructure and scalability, was appealing for its cost-effectiveness and quickly gained traction among many businesses. Since then, it has grown into the world’s largest cloud computing service, making up around 30% of the global market.

However, the popularity of AWS also makes it vulnerable. Normally, a single disruption in a company’s service isn’t a big deal. However, because the function of so many other companies depends on this one service, issues such as these are catastrophic. This reflects exactly what happened; when AWS went down for even a day, the effects were immediately noticeable.

The outage originated in US-EAST-1; this region is the oldest and most widely used in AWS, which also makes it the most disastrous in case of a failure. The issue was a problematic software update which cascaded into a bigger failure within AWS’ core infrastructure, preventing traffic from getting through their servers.

This isn’t to say that incidents like this outage are unprecedented for AWS. In fact, the outage this October is the third major malfunction in the US-EAST-1 region alone since 2020. If anything, they are quite familiar with service breaking down by this point. What’s more, the cause is due to regular company functions like software updates, rather than some form of cyberattack. Considering how many people around the world rely on AWS-dependent services, it’s quite concerning to see that outages can be this disastrous and yet remain this common.