How an AWS database hiccup froze the web

On 20 October 2025, a major internet outage knocked dozens of popular websites and apps offline, leaving users worldwide unable to log in, pay, or use familiar services for several hours. Amazon Web Services (AWS) — the cloud company that hosts whole stacks for many businesses reported the incident and engineers raced to restore normal operations.

In plain terms: many online services rely on central tools that store and fetch data. One of those tools, an Amazon database called DynamoDB, started returning errors and timing out in AWS’s largest data hub in northern Virginia (US-EAST-1). AWS engineers traced the problem to how systems were resolving the DynamoDB address (DNS). Due to the apps not being able to reach that database endpoint reliably, the failure cascaded: services that needed those database responses either crashed or became painfully slow.

The outage hit a long list of familiar names. Snapchat, Reddit, Roblox, Fortnite, Duolingo and Perplexity all reported problems. Amazon’s own retail and smart-home services stumbled. In the UK, users found banking websites, mobile networks and the HM Revenue and Customs site difficult or impossible to access. Downdetector, which collects user reports, registered huge spikes across many platforms as people tried and failed to use online services.

Affected services and platforms: 

• Snapchat — one of the first major consumer apps to report trouble.

• Fortnite — the gaming platform went down amid the cloud fault.

• Roblox — another popular gaming service experiencing outages.

• Duolingo — reported service issues during the disruption.

• Perplexity (AI startup) — explicitly attributed the outage to AWS.

• Venmo — payment app that faced connectivity/trouble due to AWS issues.

• Robinhood — trading app that also flagged problems during the broader outage.

• Slack — workplace communication tool among those impacted.

• Signal — messaging app confirmed trouble tied to AWS’s fault. 

• HM Revenue & Customs (UK) — government portal among services reporting issues.

• Various UK banks including Lloyds Bank, Halifax and Bank of Scotland reported interruptions.

• Ring smart doorbell service (owned by Amazon) had outages.

• Wordle — popular word-game showed signs of service trouble.

• A wide range of other apps and platforms: according to sources, platforms like Canva, Whatnot, Rainbow Six Siege, Life360, the McDonald’s App, and others all experienced interruptions.

• AWS internal services themselves: certain AWS core services like Amazon DynamoDB (its managed NoSQL database) were explicitly listed as “impacted”. 

• Additional AWS services listed as impacted: AWS Lambda, Amazon EC2 (Elastic Compute Cloud), Amazon CloudFront, Amazon S3, Amazon API Gateway, AWS IAM Identity Center, Amazon Simple Queue Service (SQS) and Amazon Kinesis Data Streams.

Security teams and reporters found no evidence of a cyberattack behind Monday’s internet disruption. They concluded the outage came from a technical fault inside Amazon’s own cloud network. AWS confirmed it was suffering from "increased error rates and latencies" for multiple services. It said the potential cause of the technical hitch was related to "DNS resolution of the DynamoDB API endpoint in US-EAST-1".

AWS engineers launched several recovery efforts, rerouting traffic and clearing backlogs of failed requests to restore stability. Within hours, apps and websites started returning online as Amazon’s systems recovered across affected regions.

This outage shows exactly how much daily life depends on a handful of cloud providers. When a central cloud service stumbles, it can drag many unrelated apps down at once. That means you might not be able to shop, access work tools, file government forms or see social feeds, all because one critical service hiccuped. For businesses, the fallout translates into lost sales, missed messages and frantic rerouting to backups. For consumers, the disruption usually feels like an inconvenience, but it can temporarily block access to essential services such as banking and tax filing.

What to do next: if you find a service down, check its official status page or trusted news outlets for updates and avoid repeatedly re-entering payment details. For organisations, the episode is a reminder to build redundancy across cloud providers and geographic regions so a single outage has less effect. AWS and affected companies typically publish detailed post-mortems after such incidents explaining what went wrong and how they plan to prevent a repeat.

The incident exposed how deeply the global web relies on Amazon’s infrastructure. Regulators and governments also monitor these outages closely, particularly when banks and public services feel the effects. For now, customer funds remain safe, and the cloud industry and its customers will be watching closely for lessons from this disruption. Engineers and analysts now expect AWS to tighten its regional redundancy and DNS failover systems to prevent future single-point failures. For companies building in the cloud, the message is clear: diversify or risk going dark when your provider does