Amazon Web Services·

AWS us-east-1 outage: an empty DNS record takes down DynamoDB and much of the region

A race condition in DynamoDB's DNS automation erased the service's address in AWS's largest region, and the failure spread to EC2, load balancers, Lambda and many other services for about 15 hours.

ResolvedEnded Oct 20, 2025
Security:
No determination
Started
Oct 20, 2025, 06:48 UTC
All AWS services recovered
Oct 20, 2025, 21:20 UTC
Duration
14 h 32 min
Full recovery
Complete
Geographic scope
United States (us-east-1)
Last verified
1 source

What happened

Systems and services affected
Amazon DynamoDBAmazon EC2Network Load BalancerAWS LambdaAmazon ECS / EKS / FargateAWS STS and IAM sign-inAmazon ConnectAmazon Redshift
Kind of failure
Software defect
Sector
Cloud infrastructure
Provider report
Published
Status and recovery
Resolved. AWS says the event ended at 2:20 PM PDT on October 20.[1]

AWS gives times in Pacific time: 11:48 PM PDT October 19 to 2:20 PM PDT October 20.

What caused it?

Root cause: Confirmed

A latent race condition in DynamoDB's automated DNS management left the regional endpoint dynamodb.us-east-1.amazonaws.com with an empty DNS record that the automation could not repair.[1]

Provider report: Published. AWS published a post-event summary. The page does not show its publication date.[1]

Was it malicious?

No determination

AWS's summary attributes the outage to a defect in its own automation. It does not mention an attack.[1]

How the failure happened

  1. 1TriggerLatent race condition
  2. 2FailureEmpty DNS record
  3. 3SpreadEC2 launches and load balancers failed
  4. Affected8 services and systems

Evidence for each step

  1. 1
    Latent race condition. Between DNS Enactors in DynamoDB's automated DNS management[1]
    Starting point.
  2. 2
    Empty DNS record. For dynamodb.us-east-1.amazonaws.com, which the automation could not repair[1]
    Led from step 1: Stated in a provider report or official record.
  3. 3
    EC2 launches and load balancers failed. EC2 could not re-establish leases; Network Load Balancer health checks removed healthy capacity[1]
    Led from step 2: Stated in a provider report or official record.

Only steps the sources state are drawn. Affected services are listed under What happened.

Explanation

Both versions summarize the sourced claims on this page. Our own interpretation appears only under Independent analysis.

Plain language: what happened and why people were affected.

Many apps and websites that run on Amazon Web Services stopped working for much of October 20, 2025.

AWS's database service, DynamoDB, has a web address that other systems look up to find it. An automated system that keeps that address up to date had a timing bug: an old update overwrote a newer one, and then a clean-up step deleted it. The address pointed nowhere.

Many other AWS services depend on DynamoDB, so they failed too. Fixing the address took about three hours; untangling the knock-on effects took most of the day.

Confirmed and unknown

Confirmed

  • Confirmed
    A race condition between DNS Enactors left the DynamoDB regional endpoint with an empty DNS record.[1]
    Said by: AWS · Evidence: Official
  • Confirmed
    DynamoDB API errors lasted from 06:48 to 09:40 UTC; the wider disruption ran until 21:20 UTC.[1]
    Said by: AWS · Evidence: Official
  • Confirmed
    EC2 instance launches failed from 09:25 to 20:50 UTC because DWFM could not re-establish leases and network state lagged.[1]
    Said by: AWS · Evidence: Official
  • Confirmed
    Network Load Balancer health-check failover removed healthy capacity and caused connection errors from 12:30 to 21:09 UTC.[1]
    Said by: AWS · Evidence: Official

Still unknown

No open questions: the provider has published a full account.

Timeline

Times in UTC. Dates without a time mean the sources give no time. Only events a source documents are listed.

  1. First observedOct 20, 2025, 06:48 UTC
  2. AcknowledgedNot documented
  3. MitigationOct 20, 2025, 09:40 UTC
  4. RecoveryNot documented
  5. ResolvedOct 20, 2025, 21:20 UTC
  6. Incident reportPublished, date not shown
Oct 20, 2025, 06:48 UTCOct 20, 2025, 21:20 UTC
  1. First observed
    DynamoDB endpoint loses its DNS record; API errors begin in us-east-1.[1]
  2. Mitigation
    DynamoDB DNS restored; DynamoDB errors end.[1]
  3. Engineers throttle work and restart EC2 DWFM hosts.[1]
  4. Network Load Balancer connection errors begin as health checks flap.[1]
  5. AWS disables NLB automatic health-check failover.[1]
  6. EC2 network state propagation returns to normal.[1]
  7. Resolved
    All AWS services recovered.[1]
  8. Start to: All AWS services recovered14 h 32 min

Related incidents

Links stated by providers or officials carry their evidence label. “Similar failure” is a comparison of documented causes, not a causal link.

Evidence and sources

Grouped by type of evidence. Numbers match the citations above.

Primary: provider and official sources

Published by the provider or agency itself: incident reports, status records, advisories, statements.

  1. 1.
    Summary of the Amazon DynamoDB Service Disruption in the Northern Virginia (US-EAST-1) Region (opens in a new tab)
    Amazon Web Services · Provider report · Publication date not recorded · Accessed

Last verified against these sources on Nov 14, 2026.