Summary: Cloudflare’s R2 object storage experienced an outage of 1 hour and 7 minutes due to a credential misconfiguration, leading to 100% write and 35% read failures worldwide. The issue arose when new credentials were mistakenly deployed to a development environment instead of production, leaving the service without valid authentication. Cloudflare is implementing improved logging, verification, and dual validation for high-impact actions to prevent future incidents.
Affected: Cloudflare R2 Object Storage
Keypoints :
- Outage led to 100% write failures and 35% read failures for R2, with some cached content remaining accessible.
- Incident caused degraded service levels for various other Cloudflare services, including image delivery and email security.
- Cloudflare is enhancing procedures for credential management to mitigate future risks, including automated deployment tooling and stricter validation processes.