• 2 min read
AWS CloudFront outage knocks Hugging Face offline
An AWS CloudFront issue started at 0145 PDT, serving 5xx errors to customers using VPC Origins and disrupting services including Hugging Face and the UK National Lottery.
Image: The Register
An AWS outage tied to CloudFront began at 0145 PDT (0945 UTC), sending 5xx errors to websites and apps that rely on the service and knocking a string of online services offline across multiple regions.
According to AWS, the incident affects CloudFront customers using VPC Origins, a newer feature that lets customers serve applications behind private load balancers through CloudFront without exposing back-end infrastructure to the public internet. AWS said customers using other origin types are not affected, and recommended that those who can do without VPC Origins temporarily switch origin types while engineers work on a fix.
On its service status page, AWS said:
“We are experiencing increased 5xx errors for CloudFront customers utilizing VPC Origins connectivity. Our engineers are engaged and are actively working to mitigate impact.”
At 03:18 PDT (118BST, 1018 UTC), the company added that it believed the root cause was tied to a packet processing subsystem that routes requests from CloudFront edge locations to resources inside customer VPCs. AWS said it would provide another update within the hour.
Users hitting affected sites were shown an error message saying the service could not connect to the server and suggesting there might be too much traffic or a configuration issue.
Recommended reading
Shadow AI grows when workplace tools get in the way
Among the early visible casualties was Hugging Face, which said its service was unavailable “from most regions in the world” because of the AWS outage while it worked on mitigation. The UK’s National Lottery also said on X that players could not access its website or mobile app due to what it described as a wider AWS outage.
The disruption also hit gaming services. Fallout 76 players posted on Reddit about access problems, while other AWS customers reported the same pattern: CloudFront distributions returning 5xx errors while other AWS services appeared to operate normally.
AWS had not disclosed the cause at first and did not immediately respond to The Register’s questions, though its later status update pointed to the packet-processing system handling VPC routing.
Enterprise Editor
Marcus follows the money. He covers enterprise software, cloud architecture, and the tectonic shifts in Big Tech strategy. He translates dense earnings calls and complex M&A activity into actionable insights about where the industry is actually heading. If a tech giant makes a silent pivot, Marcus is usually the first to notice.
via The Register
