Cloudflare Outage History

The issue affected Cloudflare's tiered cache in Ashburn, VA, causing elevated response times for some requests. The incident began at Jun 12, 2025 12:46 AM UTC with the team investigating the latency, and the status was raised to a warning at 12:50 AM UTC. The problem was resolved by 1:40 AM UTC when operation returned to normal. The provider issued 2 official updates during the incident to communicate the investigation and the resolution. The impact was regional to Ashburn and did not appear to be a global outage.

Cloudflare reported degraded WARP connectivity and impact to Cloudflare Zero Trust users beginning in the early hours of Jun 11, 2025 4:31 AM UTC. The issue affected WARP connectivity and potentially degraded Internet experience for affected users. Later, a regional symptom emerged with increased errors for Durable Objects in the Asia-Pacific (APAC) region, suggesting a broader underlying fault. The incident spanned roughly 20 hours, with service gradually returning to normal as updates progressed. The outage was tracked across multiple official updates and involved changes to the WARP component and APAC Durable Objects status. The incident began with the first status update at Jun 11, 2025 4:35 AM UTC and progressed to a regional alert at Jun 11, 2025 2:45 PM UTC, before the final resolution at Jun 12, 2025 12:35 AM UTC.

Cloudflare conducted scheduled maintenance in the Zurich region (ZRH) starting at 1:04 AM UTC. The operation briefly placed traffic in a degraded state beginning at 1:05 AM UTC and completed by 4:05 AM UTC when service returned to normal. The provider issued two status updates during the window, and the impact was localized to the Zurich region with no global outage. The maintenance window lasted about 3 hours.

Cloudflare conducted scheduled maintenance at the ORD Chicago data center, causing a brief outage window for traffic routed through that location. The maintenance window ran from 8:00 AM UTC to 10:00 AM UTC. Cloudflare posted two official updates during the window, and ORD Chicago was the affected component. Services routed through ORD returned to normal by 10:00 AM UTC, with the impact localized to that data center.

Cloudflare carried out a scheduled maintenance window in the ORD (Chicago) region beginning around 8:00 AM UTC. The incident was listed as in progress with an update stating that maintenance is underway and updates would be provided as necessary.

The maintenance ran for about 2 hours and 5 minutes, with services in the ORD region returning to normal by 10:05 AM UTC. The ORD region was the primary area affected during this window.

Cloudflare experienced a brief disruption affecting the Workers AI component. Users attempting AI inference with the @cf/meta/llama-3.3-70b-instruct-fp8-fast model could have seen increased errors. The incident began at Jun 7, 2025 1:50 PM UTC and was resolved by Jun 7, 2025 2:25 PM UTC, lasting about 35 minutes.

Cloudflare published status updates acknowledging the issue and detailing mitigation steps. The incident involved the Cloudflare Sites and Services / Workers AI component, and the provider issued 2 status updates during the incident.

Cloudflare reported possible network congestion in Hong Kong beginning at 7:40 AM UTC, with the Asia / Hong Kong (HKG) component affected. Users in the Hong Kong region may have experienced connectivity issues or slower network performance while Cloudflare investigated the cause. No user reports were observed during the incident, and the issue appeared localized to the Hong Kong region with no broader global impact indicated.

Three official updates were published during the incident: the initial alert at 7:40 AM UTC, a follow-up at 7:45 AM UTC confirming ongoing investigation, and the resolution at 12:55 PM UTC. The outage lasted roughly 5 hours and 15 minutes, ending when the status returned to up. The impact was confined to the Asia / Hong Kong (HKG) component, with no other regions noted as affected.

Cloudflare's CDN/Cache experienced a surge of 502 errors that degraded content delivery and caused errors when loading resources for sites using Cloudflare. The incident began at Jun 6, 2025 4:50 PM UTC and persisted until Jun 6, 2025 10:10 PM UTC; a fix was implemented by Jun 6, 2025 8:02 PM UTC and monitoring followed as Cloudflare worked to restore normal operation.

Cloudflare carried out scheduled maintenance at CNF Belo Horizonte in Brazil. The maintenance window began at Jun 6, 2025 8:15 AM UTC and the status was updated to warn as work progressed at Jun 6, 2025 8:20 AM UTC, indicating possible impact to Cloudflare services during the window. The maintenance concluded with the status returning to normal at Jun 6, 2025 12:50 PM UTC.

Cloudflare experienced a two-pronged incident on Jun 5, 2025 9:00 PM UTC. An issue with newly deployed Snippets at the edge prompted an initial warning, though the control plane, dashboard, and API remained unaffected. The issue escalated when the Realtime STUN/TURN service went down at 10:40 PM UTC, impacting real-time connectivity for affected users. A Snippets fix was implemented and monitoring began by 11:01 PM UTC, and the Realtime TURN service recovered by 11:45 PM UTC. Cloudflare issued multiple status updates during the incident, and user reports were isolated, including observations from Samara, Russia. The outage lasted about 1 hour and 5 minutes.