AWS had a 2h 18m connectivity incident in US-EAST-1 (N. Virginia) that broke authentication through IAM Identity Center, blocked AWS Sign Up, and disrupted services that depend on Identity Center logins, including Amazon Connect. This was a fault on AWS's side, not yours.
Started
12:38 UTC
Duration
Lasted 2h 18m
Source
IsDown
Next time
AWS (Amazon Web Services) is degraded. Is your product working?
An incident on their side does not always mean an outage on yours, and the only way to know is to already be watching your own endpoints. Set that up in under a minute and you will be notified the next time your own site is affected. And there will be a next time.
50 monitors for your own websites. No card.
Affected components: Multiple services, Multiple services (N. Virginia), AWS IAM Identity Center, AWS IAM Identity Center (N. Virginia), AWS Sign Up, Amazon Connect, Amazon Connect (N. Virginia), Amazon CloudWatch, Amazon CloudWatch (N. Virginia), Amazon VPC Lattice, Amazon VPC Lattice (N. Virginia), AWS IoT Device Management, AWS IoT Device Management (N. Virginia), AWS NAT Gateway, AWS NAT Gateway (N. Virginia), Amazon OpenSearch Service, Amazon OpenSearch Service (N. Virginia), AWS Step Functions, AWS Step Functions (N. Virginia), Amazon Elastic Load Balancing, Amazon Elastic Load Balancing (N. Virginia)
What happened?
AWS logged a single incident starting at 12:38 UTC, titled "Increased Connectivity issues in US-EAST-1 - N. Virginia", and closed it 2 hours and 18 minutes later. The first update said AWS was investigating connectivity issues in US-EAST-1 and flagged that IAM Identity Center sign-in and AWS Sign Up were impacted. By 13:15 AWS confirmed authentication errors for anyone signing in through Identity Center or through services that rely on it, and stated plainly that there were no workarounds available to affected customers. Twenty-one components were listed as affected across the day, covering a wide spread including Amazon CloudWatch, Amazon VPC Lattice, AWS IoT Device Management and Amazon Connect, each duplicated for the N. Virginia region specifically. At 13:22 AWS reported early signs of recovery for authentication and sign-up. By 14:23 AWS said most services had recovered and that the IAM Identity Center authentication issue affecting Kiro, Connect and other dependent services was fully mitigated, but it disclosed a separate, narrower problem: a subset of Network Load Balancers in a single cell, in one Availability Zone (use1-az1) of US-EAST-1, was still affected. For that NLB issue AWS recommended customers needing immediate resolution create a new NLB with targets spread across multiple Availability Zones. The final update at 14:32 reported significant signs of recovery with AWS continuing to monitor for full recovery. No cause was given for either the authentication failures or the NLB problem at any point, and AWS has not filed anything further to explain root cause. No user reports were filed during this window, so the only account of impact comes from AWS's own updates.
Learning
UptimeRobot's own probes against DynamoDB, S3 and SQS endpoints in us-east-1, eu-west-1 and ap-southeast-1 stayed at 100% uptime across 24 hours, 7 days and 30 days, and this incident was never registered against those checks. That is not a contradiction: the fault sat in authentication through IAM Identity Center and in a single Availability Zone cell behind some Network Load Balancers, not in the storage, database or queueing endpoints being probed. A service can fail for anyone trying to log in while the data layer underneath keeps answering every request normally, because sign-in and data access are different failure paths with different blast radii. If your product depends on IAM Identity Center for authentication, your own health checks on your application would likely have shown errors at the same time AWS's status page did, since that failure was close to the surface. But if your monitoring only watches your own endpoints, a problem confined to AWS's identity layer or a single Availability Zone's load balancers can pass through invisible to you until users start failing to log in. Your monitoring covers your half, UptimeRobot now watches the provider's half too, so next time you know which side broke without guessing.