A major Amazon Web Services (AWS) outage obstructed access to essential apps and platforms like Schoology, College Board as well as popular social media and gaming sites nationwide for nearly 15 hours on Oct. 20, starting in the early morning.

The incident highlights that just a few major providers are responsible for nearly all internet infrastructure. AWS controls more than 39% of the global cloud computing market share. This concentration means that when Amazon’s systems fail, the ripple effects can simultaneously impact millions of users.

Economics teacher Dean Lizardo encountered issues while trying to update an assignment on AP Classroom. 

“We may not think about it, but many websites rely on Amazon for storing their data or using their computing power,” Lizardo said. “While I personally didn’t have such a huge impact, because it was easy for me to pivot, there are some businesses that rely heavily on it. If so much of our economy relies on this single entity to provide computation or to provide storage, then that single entity actually has a lot of market power.” 

StarTree Head of Engineering Amol Mathur has spent more than a decade working with AWS, and his company depends heavily on AWS infrastructure. During the outage, StarTree encountered difficulties in launching new compute resources, resulting in a flurry of alerts and alarms across its systems. Mathur explained why AWS continues to hold such a dominant position in the cloud computing industry.

“AWS was the first major cloud provider, the company that started the idea of renting computer power and storage over the internet instead of buying your own servers,” Mathur said. “Because AWS started early and didn’t have much competition for years, it got a head start. This allowed it to gain lots of customers and build a big community of engineers who know how to use its tools. Over time, AWS learned how to run huge systems reliably and make them easier for people to use.”

For students like junior Ishani Vadali, who was unable to complete her AP Calculus assignment before its deadline because College Board’s website went down, the outage revealed how technology-based education often lacks a backup plan. 

“It’s okay to assign work through those [platforms], but if this situation ever comes up, teachers should have a way to print out the work that you have online,” Ishani said. “There needs to be a solution to quickly get the work into the hands of students even if they don’t have access to it through the platform.”  

The outage stemmed from a Domain Name System (DNS) error in AWS in Northern Virginia, referred to as the US-EAST-1 region, and caused a chain reaction. DNS is the internet’s directory system, translating website names into language that computers can understand. When this system failed, applications lost the ability to locate and retrieve user data stored on Amazon’s servers. 

While AWS reported mitigating the issue by 2:24 a.m. PDT, platforms continued to experience interruptions throughout the day due to complications with launching the new virtual servers that power many web applications. Thus, even after the initial fix, many services remained unstable or inaccessible for hours after. Full service was not restored until late afternoon on Oct. 20, marking one of AWS’s most significant disruptions in recent years.

“Even companies that don’t use AWS can learn a lot from this outage,” Mathur said. “The main lesson is that no system is perfect, and even the biggest tech companies can have unexpected failures. That means every organization should be prepared for something to go wrong, no matter how reliable their systems seem.”