On 19-20 October 2025, Amazon Web Services had a major outage in its us-east-1 region that lasted about 14.5 hours. The cause was a race condition in DynamoDB's automated DNS management, which left the regional endpoint with an empty DNS record. The failure spread to EC2, load balancers, Lambda and many other AWS services, and to the countless websites and apps that run on them.
What is going on
DNS is the internet's address book. It turns a name like dynamodb.us-east-1.amazonaws.com into the numeric address a computer can actually connect to.
AWS updates those entries automatically. A race condition is what happens when two automated jobs act at the same time and step on each other's work, in the wrong order.
In this case the result was an empty DNS record for the DynamoDB regional endpoint. The address book page was blank, so services could not find DynamoDB, and the problem cascaded outward to other AWS services that depend on it.
Who is affected
Anything hosted in or dependent on AWS us-east-1 was at risk during the event. That includes websites on EC2, applications behind AWS load balancers, serverless functions on Lambda and data stored in DynamoDB.
It also includes small businesses that never touch AWS directly. If your booking tool, payment provider, chat widget or SaaS dashboard runs in that region, your site can break even when your own server is fine.
What to do now
- List your dependencies. Write down every third-party service your site loads or calls, and find out which region each one runs in.
- Avoid a single region where you can. For critical parts, use a multi-region setup or a simple fallback, such as a static holding page served from another provider.
- Set sensible timeouts and retries. A slow third-party call should fail fast rather than freeze your whole page, and retries should back off instead of hammering a broken service.
- Host your status page somewhere else. If your status page sits on the same infrastructure as your site, it goes down with it.
- Check your backups restore elsewhere. Confirm you can bring data back up with a different provider or region if you need to.
How to check your own site
An outage is not a vulnerability, but the same review is a good moment to look at your site's basics. The free Website Security Check spots outdated CMS versions, known-vulnerable plugins, missing security headers, weak TLS and email-authentication gaps, so you know what else needs attention while you map your hosting dependencies.
Source: AWS service event summary, October 2025.
Comments
No comments yet.