Overview
- The disruption began around 3:11 a.m. ET in AWS's US-East-1 region and was tied to a DNS-resolution issue linked to DynamoDB after an API update, which cascaded to other services such as EC2, SQS and Connect.
- AWS reported the DNS issue was fully mitigated by 6:35 a.m. ET with most operations succeeding, but it warned of throttling as engineers cleared backlogs and later rate-limited new EC2 instance launches during recovery.
- By the evening of Oct. 20, AWS said all services had returned to normal operations, with some services finishing backlog processing over the following hours.
- Outage reports spanned Amazon services and third-party platforms including Snapchat, Canva, Coinbase, Roblox, Fortnite and banks such as Lloyds, Halifax and Bank of Scotland, with UK tax authority HMRC also affected and airport kiosk issues reported at New York’s LaGuardia.
- Coverage and expert commentary indicated no sign of a cyberattack, describing a technical fault that underscored global dependence on a few cloud providers and the need for stronger redundancy and contingency planning.