Amazon Web Services (AWS) has identified the root cause of this week’s hours-long outage, which affected thousands of websites and applications. The issue was caused by a bug in automation software that managed DNS records for DynamoDB, a database system used by AWS customers to store data.
The problem arose from an empty DNS record for the US-East-1 data center region. Since the bug failed to automatically repair itself, manual operator intervention was required to correct it. As a result, thousands of sites and applications hosted on AWS experienced outages, including popular platforms like Signal, Snapchat, and Roblox.
Downdetector reported over 8.1 million user complaints about the outage, with services such as banking websites and the Ring doorbell company also affected. Even smart bed companies like Eight Sleep were impacted, leaving customers unable to control their beds due to connectivity issues.
AWS has since disabled the automated DNS management system worldwide while it works to fix the issue and adds extra protections. The company’s CEO has apologized for the inconvenience and rolled out an update to its services that will allow users to control critical functions via Bluetooth in case of another outage.
Source: https://www.theguardian.com/technology/2025/oct/24/amazon-reveals-cause-of-aws-outage