Amazon Web Services Outage on October 20, 2025: Causes and Consequences
20 october 2025 в 20:28
On October 20, 2025, everything from your favorite apps to major banking services faced issues due to an unexpected outage in Amazon Web Services. Over 4 million users reported disruptions, and a warning from security threat veteran Rayf Pilling — «any problem can cause serious trouble» — indicated the scale of the outage.
Currently, AWS is tracking the aftermath and restoring services, here is a closer look at what went wrong, why, and how the internet is returning to normal.
Amazon Web Services (AWS) is the cloud computing division of Amazon, offering on-demand services such as data storage, computing power, and networking solutions for businesses and applications worldwide.
The outage that occurred in AWS on October 20, 2025, began around 3 am Eastern Time in its US-EAST-1 region (Northern Virginia), where the company reported a surge of «errors and delays» affecting over 80 internal services.
AWS ultimately traced the outage to a DNS resolution failure related to the DynamoDB database endpoint. AWS stated that the root of the problem was infrastructural, not a cyberattack, although the company did not provide full public information.
According to the latest information, AWS reports that most affected services are now operational again. The company stated that it sees «significant signs of recovery» and that «most requests should now be successfully executed». Some users may still experience delays or minor disruptions, but major disruptions seem to have been mostly resolved.
At 6:35 pm Eastern Time, AWS shared a more detailed update, saying: «The main issue with DNS has been fully resolved, and most AWS Service operations are now successfully executing. Some requests may be limited as we work on a complete resolution. Additionally, some services continue to operate with event backlog, such as Cloudtrail and Lambda. While most operations have been restored, requests to launch new EC2 instances (or services that launch EC2 instances, such as ECS) in the US-EAST-1 region continue to experience elevated error rates. We continue to work on a complete resolution».
Although the main issue has been resolved and service restoration is underway, AWS warned that full recovery may take additional hours as backlogged requests are processed and systems stabilize.
Since many applications and websites rely on interconnected cloud services, users may continue to experience partial issues until all infrastructure levels return to normal
Currently, AWS is tracking the aftermath and restoring services, here is a closer look at what went wrong, why, and how the internet is returning to normal.
Amazon Web Services (AWS) is the cloud computing division of Amazon, offering on-demand services such as data storage, computing power, and networking solutions for businesses and applications worldwide.
The outage that occurred in AWS on October 20, 2025, began around 3 am Eastern Time in its US-EAST-1 region (Northern Virginia), where the company reported a surge of «errors and delays» affecting over 80 internal services.
AWS ultimately traced the outage to a DNS resolution failure related to the DynamoDB database endpoint. AWS stated that the root of the problem was infrastructural, not a cyberattack, although the company did not provide full public information.
According to the latest information, AWS reports that most affected services are now operational again. The company stated that it sees «significant signs of recovery» and that «most requests should now be successfully executed». Some users may still experience delays or minor disruptions, but major disruptions seem to have been mostly resolved.
At 6:35 pm Eastern Time, AWS shared a more detailed update, saying: «The main issue with DNS has been fully resolved, and most AWS Service operations are now successfully executing. Some requests may be limited as we work on a complete resolution. Additionally, some services continue to operate with event backlog, such as Cloudtrail and Lambda. While most operations have been restored, requests to launch new EC2 instances (or services that launch EC2 instances, such as ECS) in the US-EAST-1 region continue to experience elevated error rates. We continue to work on a complete resolution».
Although the main issue has been resolved and service restoration is underway, AWS warned that full recovery may take additional hours as backlogged requests are processed and systems stabilize.
Since many applications and websites rely on interconnected cloud services, users may continue to experience partial issues until all infrastructure levels return to normal
© Puhova Marina
