Overview
- CloudFront began returning elevated 5xx server errors for distributions that use VPC Origins after 12:45 AM PDT, and customers using other origin types were not affected.
- AWS identified a likely fault in a packet‑processing routing subsystem that directs requests from CloudFront edge locations to customer VPCs and tied the problem to Availability Zone euc1-az2 in Frankfurt.
- The disruption produced global user impacts because CloudFront’s edge network could not reach the affected private origins, causing websites and apps to show 500–504 errors even when backends remained healthy.
- AWS has advised customers who can avoid VPC Origins to temporarily change their origin type as a workaround while engineers test a mitigation that will be rolled out in phased stages.
- Outage trackers and social reports showed a sharp spike in user complaints and the incident recalls a large October AWS outage, underlining how many consumer and business services depend on a small number of cloud infrastructure providers.