An AWS outage takes down applications and businesses worldwide

What happened

On 20 October 2025 AWS suffered an outage in its US-EAST-1 region (Northern Virginia). According to AWS, the root cause was a DNS resolution problem affecting the DynamoDB database service, which cascaded to other services in the region for hours.

Consumer apps, games, banking and business tools worldwide were affected, including some whose users were outside the US, because they relied on services hosted in that region.

What it means for a business in Spain

Many companies don't buy AWS directly but use software that depends on it. The relevant question isn't how big the provider is, but what happens to the business if a specific region fails.

Designing to tolerate that failure costs money, so it should be decided per application, based on the real cost of downtime.

What to review

Useful questions for each critical application:

• Which region and provider it runs on, directly or indirectly.
• Whether a copy of the data exists outside that provider.
• How long it can be down without serious harm.
• What manual or secondary alternative exists during recovery.

Frequently Asked Questions

Was it an attack?

No. AWS attributed the outage to an internal fault related to DynamoDB DNS resolution in US-EAST-1.

Does being in the cloud guarantee high availability?

Not on its own. Availability depends on application design: regions, replicas and dependencies.

Does combining public cloud with a private datacenter make sense?

Often, yes. It allows backups or critical services to live outside a single provider.