Microsoft Cloud IT Pro Podcast

Episode 414 – When the Cloud Falls: Understanding the AWS and Azure Outages of October 2025

6 snips
Nov 6, 2025
In October 2025, both AWS and Azure faced significant disruptions due to DNS outages. The hosts delve into the root causes, including complex microservices and configuration changes that amplified the failures. They discuss how even multi-cloud strategies, exemplified by Starbucks, weren't foolproof against outages. Insights into Azure’s internal service changes and the technical issues leading to AWS failures are highlighted. The conversation wraps up with recommendations for IT pros on improving disaster recovery plans and communication protocols.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
INSIGHT

Small Internal State Failures Cascade Widely

  • Major cloud outages often stem from complex internal state and DNS interactions rather than simple network failures.
  • These failures cascade widely because many services depend on shared cloud components like DNS and global load balancers.
ANECDOTE

Meme About DNS Engineers Switching Clouds

  • Ben shared a meme joking about someone moving from AWS to Azure after DNS/IPv6 misconfigurations caused outages.
  • The joke highlights how DNS misconfigurations have become a recurring theme across cloud providers.
INSIGHT

Race Conditions Threaten Distributed Configs

  • Race conditions and desynchronized workers can apply stale configs and override correct ones, causing cascading failures.
  • Hyper-scale microservice architectures make these subtle timing issues systemic risks across services.
Get the Snipd Podcast app to discover more snips from this episode
Get the app