These questions are now top of mind after a recent and unprecedented AWS outage. It shines a spotlight on our reliance around the world on just a few dominant technology firms. This week’s Container Security incident took down thousands of websites and applications. As a result, countless businesses and residents lost access to essential services. Dr. Suelette Dreyfus, a University of Melbourne lecturer in computing and information systems, called attention to the fact that the outages have critical implications. As she explained, they are a showcase of what happens when we build single points of failure into our internet infrastructure.
On Thursday, AWS released a long and informed postmortem explaining how the outage happened. The entire progression of events that led to the final outcome of service outages for thousands of customers became clear after reading the report. This ongoing incident serves as a reminder of the significant impact that technical outages at large cloud providers can have. These challenges extend throughout the entire internet ecosystem.
Matteo Franceschetti, the chief executive of AWS, went on the X social media platform to address the outage publicly and to apologize to customers. He acknowledged the disruption the incident created. He promised clients something would be done so it wouldn’t happen again.
Dr. Dreyfus emphasized that this increasing reliance on AWS and other large cloud providers poses serious threats to the resilience of the internet itself. She stated, “That single point isn’t just AWS – they’re the biggest cloud provider with 30% or so of the market – but rather the cloud as a whole, which is basically just three companies.” This short statement hints at a more ominous trend, where a few corporations and foreign governments decide and control the essential infrastructure.
The unintended consequence of the incident was that AWS improved its services significantly as a result. The company took customer concerns about possible future outages to heart. In retaliation, they released a patch that allows users to maintain basic functions on their devices such as smart beds by using Bluetooth during outages. This is an important new step toward preventing the disruptive effects of future outages.
Dr. Dreyfus further elaborated on the implications of reliance on major cloud providers, stating, “The internet was designed to be resilient. Many other channels existed for routing around problems or attacks, but we’ve lost some of that resilience by becoming so dependent on a handful of giant tech companies to provide not just data storage but house data services.” This realization is prompting a re-evaluation of how companies and consumers interact with cloud technologies.
The outage has wider repercussions for any sector that depends upon cloud services, including finance and health care. Particularly, AWS’s DynamoDB stores hundreds of thousands of DNS records, so its stability is very important for online services.