AWS restores services after overnight DNS failure in US-EAST-1 region

Amazon Web Services (AWS) has announced that all systems have been fully restored following a major DNS failure in the US-EAST-1 region, which caused widespread disruption to several key platforms overnight.

The incident began at 11:49pm (PDT) on October 19 and lasted until 2:24am (PDT) on October 20, affecting Amazon.com, AWS Support, and multiple third-party services built on its infrastructure.

Cause and impact

AWS confirmed that the disruption stemmed from a DNS resolution failure impacting DynamoDB service endpoints in the US-EAST-1 region. The issue prevented systems from resolving domain names, interrupting critical operations across its network.

Engineers identified the root cause at 12:26am (PDT) on October 20 and immediately began implementing fixes.

Recovery timeline

By 2:24am (PDT), AWS had resolved the DNS malfunction, allowing most services to recover gradually. However, certain internal systems remained unstable, prompting temporary feature restrictions, including limits on launching new EC2 instances, to stabilise operations.

Restoration efforts continued through the morning, and by 12:28pm (PDT), most customers were able to access their AWS services as normal. The company then progressively lifted EC2 restrictions while addressing minor residual issues.

At 3:01pm (PDT), AWS confirmed that all systems were fully operational, marking the complete recovery of its cloud infrastructure after more than 15 hours of disruption affecting a wide range of global users.