Cloudflare Outage History

Cloudflare experienced an issue affecting Workers invocations where customers encountered 1003/403 errors when making fetch() calls inside Workers. The problem impacted Cloudflare Sites and Services along with the Workers platform for 7.8 hours. A fix was identified and implemented, with most customers seeing resolution and a continued rollout for remaining affected users.

Cloudflare experienced network performance issues in Phoenix, AZ that caused errors and degraded performance for customers whose traffic was routed through that location. The incident affected Cloudflare Sites and Services in the Phoenix area for 22 minutes before being resolved.

Cloudflare experienced elevated error rates for R2 storage buckets in the APAC region, affecting customer access to their stored data and associated sites and services. The incident lasted 1.7 hours while Cloudflare investigated and resolved the connectivity issues. The problem impacted customers attempting to access their R2 buckets during the outage period.

Cloudflare experienced increased upload failures for Workers Assets over 5.2 hours, preventing customers from successfully uploading new assets. The serving of existing Workers Assets traffic remained unaffected throughout the incident. The issue was identified and resolved with a implemented fix, followed by monitoring to ensure stability.

Cloudflare experienced intermittent 500 errors affecting their Realtime APIs for 5.7 hours, while RealtimeKit meetings continued to function normally. The incident did not impact CDN file serving or edge security features, and Cloudflare investigated the API issues throughout the duration.

Cloudflare Pages experienced a 56-minute incident where customers encountered failures and errors when attempting to add or manage custom domains through the Pages API or dashboard. The issue was caused by problems with the custom domain management API and high latency/intermittent errors in the SSL for SaaS Custom Hostnames API. Deployed Workers and Pages projects continued serving traffic normally throughout the incident, with only custom domain management functionality being affected.

Cloudflare experienced an issue that caused increased 502 errors for a subset of customers, affecting their Sites and Services and CDN/Cache components. The problem was identified and a fix was implemented within 24 minutes, with monitoring ongoing to ensure resolution.

Cloudflare experienced an issue where Spectrum HTTP/HTTPS applications were reporting incorrect client source IP addresses and geolocation information. The problem affected Cloudflare Sites and Services and Spectrum components for 2.1 hours. Cloudflare identified the root cause, implemented a fix, and monitored the resolution.

Cloudflare experienced issues with their Workers AI service where customers encountered errors when making inference requests to the @cf/openai/gpt-oss-120b model. The incident affected Cloudflare Sites and Services along with Workers AI functionality. A fix was implemented and deployed within 1.4 hours, with Cloudflare monitoring the results to ensure resolution.

Cloudflare experienced intermittent connectivity issues with Access Browser-based RDP functionality, affecting customers' ability to reliably connect to remote desktop services through their Access platform. The incident impacted Cloudflare Sites and Services and Access components for approximately 3.4 hours. The service disruption caused transient connection problems for users attempting to access RDP sessions through Cloudflare's browser-based interface.