azure front door outage: a latent bug disrupted azure and m365
A latent bug in Azure Front Door crashed edge servers worldwide, causing timeouts and DNS failures for Azure, Microsoft 365 and sites fronted by AFD from mid-afternoon UTC until just after midnight.
- tracking id
- YKYN-BWZ
- impact window
- 15:41 UTC 29 Oct – 00:05 UTC 30 Oct 2025 (Microsoft PIR)
- root cause
- incompatible config metadata across two control-plane versions triggered a latent data-plane crash bug
- services hit
- Azure Front Door, Azure CDN, Microsoft 365, parts of Entra ID, App Service, Azure SQL Database, Azure Portal, Static Web Apps and more
what happened
- Microsoft's post-incident review (tracking ID YKYN-BWZ) puts the impact window at 15:41 UTC on 29 October 2025 to 00:05 UTC on 30 October 2025.[1]
- Customers and Microsoft services using Azure Front Door and Azure CDN saw connection timeouts and DNS resolution failures.[1]
- Root cause: a sequence of valid customer configuration changes made across two control-plane build versions produced incompatible metadata that triggered a latent bug and crashed the data plane.[1]
- Because the crash surfaced only after about five minutes, the protection system let the change advance, and the faulty configuration was saved as the 'last known good' snapshot.[1]
- Affected services included Microsoft 365, parts of Entra ID, App Service, Azure SQL Database, the Azure Portal and Azure Static Web Apps.[1]
- Engineers manually removed the bad configuration from the snapshot, deployed it from 17:40 UTC, and kept AFD configuration changes blocked until 5 November 2025.[1]
sources1
- 01Post Incident Review (PIR) – Azure Front Door – Connectivity issues across multiple regions ↗ primaryMicrosoft Azure Statusaccessed 2026-10-21