Amazon Web Services (AWS) has recovered from a widespread outage that rippled across the internet on Monday, knocking popular apps and websites offline for hours and slowing others to a crawl. Amazon attributed the incident to DNS resolution problems affecting DynamoDB endpoints in the US-EAST-1 (Northern Virginia) region, which cascaded across dependent services; the company says mitigation was completed at 09:24 UTC (12:24 EEST) on Oct. 20.
What happened
- Start & fix window: Elevated errors began 06:49 UTC (09:49 EEST) and were mitigated by 09:24 UTC (12:24 EEST), per Amazon’s post-incident update.
- Cause (per AWS): DNS resolution failures for DynamoDB regional endpoints in US-EAST-1, triggering downstream issues across multiple AWS services. Independent network observers reported the impact propagating to third-party platforms.
- No sign of a cyberattack, according to early assessments cited by reporters.
Major consumer and enterprise apps experienced outages or degraded performance, including Snapchat, Reddit, Alexa, Fortnite, Zoom, Venmo, Coinbase, and Ring, with spillovers to some UK banking services and government sites. Some platforms recovered quickly; others processed backlogs through the day.
Outage trackers logged millions of user reports across hundreds to over a thousand companies, underscoring how failures in one cloud region can ripple globally.
US-EAST-1 is among AWS’s busiest regions and a historical flashpoint for incidents. Monday’s disruption again highlights the internet’s dependence on a handful of cloud providers and the need for multi-region designs, diversified DNS, and fault-tolerant architectures to limit single-region blast radius.
Leave a Reply