Clients across several AWS accounts and regions are down at the same time. The applications work when accessed through a VPN that bypasses CloudFront, so I'm wondering if CloudFront is having an issue. Has anyone found a more specific cause?
Is CloudFront having an outage, or is it just VPC origins?
Asked by
Ruflet
Answers
VPC origins look like the common factor, rather than every CloudFront distribution being down. Reports mention failures across multiple regions and accounts, while some static HTML sites served through CloudFront still worked.
Check whether the affected distributions use VPC origins, including a private load balancer. A working non-VPC origin would help narrow it down, though it wouldn't prove that every other CloudFront feature is unaffected.
Reply to this answer
A reported AWS Health notice described increased 5xx errors for CloudFront customers using VPC origins. Check the current AWS Health status for updates, since an earlier check didn't show anything and the situation could have changed.
Reply to this answer
The VPN test is a useful clue. It suggests the application can respond through that route, but it doesn't isolate every part of the CloudFront-to-origin path. Compare the affected distributions, their origin types, and the errors being returned before changing the application or database.
Reply to this answer
Don't assume switching to a regular origin will be a quick fix. There was also a report of trouble creating a non-VPC origin, although it wasn't clear how widespread that was.
Reply to this answer
Add your answer
Answers are reviewed before publication.
Related Discussions