Cloudflare Outage History

Cloudflare experienced a broad outage spanning multiple services, including Workers API, Pages, Dashboard, and CDN/Cache, with degraded performance and potential for failed deployments. Some updates noted that edge traffic was not affected, but the control plane and deployment surfaces were impacted. The incident began on Aug 19, 2025 8:00 AM UTC with scheduled maintenance and escalated into a widespread issue affecting sites and apps relying on Cloudflare's edge network.

Cloudflare issued frequent status updates as they identified the issues and worked on fixes, with updates covering Workers, Snippets, Pages, Dashboard, CDN/Cache, and related services such as WARP, Stream, Analytics, and R2. The incident persisted for an extended period, and the status returned to up on Sep 4, 2025 1:05 PM UTC as a resolution was achieved.

User reports indicated broad geographic and service impact across regions, consistent with a large, global disruption. The outage lasted from Aug 19, 2025 8:00 AM UTC to Sep 4, 2025 1:05 PM UTC and involved multiple updates to keep customers informed.

Cloudflare carried out scheduled maintenance beginning around Aug 18, 2025 10:01 PM UTC. The status page listed the event as in progress and later transitioned to a warning at Aug 18, 2025 10:05 PM UTC. The maintenance window ended with Cloudflare returning to normal operation by Aug 19, 2025 6:05 AM UTC, roughly 8 hours after it began. The provider issued 2 status updates during the incident.

The incident affected Cloudflare WARP and Cloudflare Zero Trust services during scheduled maintenance in the Chișinău region (KIV). Users relying on WARP and Zero Trust could have experienced issues with the public API, device registration, and managed networks. The outage lasted about 1 hour, with regional maintenance in progress at the time.

Cloudflare reported an incident affecting Authenticated Origin Pulls, with intermittent HTTP 520 responses impacting requests routed through Cloudflare's CDN/Cache. The issue began around 11:42 PM UTC and the status transitioned from up to warn at 11:45 PM UTC as engineers began investigating. The impact was felt by users relying on Authenticated Origin Pulls, causing degraded origin fetches for periods of time. The disruption lasted about 11 hours and 13 minutes.

The incident persisted until 10:55 AM UTC when Cloudflare listed the service as back to normal. The Cloudflare status page showed two updates during the incident as teams worked to mitigate the issue.

Cloudflare reported a brief degradation in analytics data delivery affecting usage analytics for Durable Objects and D1. The issue began at 2:50 PM UTC and was resolved by 3:50 PM UTC, lasting about 1 hour. Cloudflare posted two status updates during the incident, documenting the investigation and restoration of analytics delivery. The impact was limited to analytics reporting for these components, with no broad service outages observed.

Cloudflare D1 experienced a brief degradation in its control plane, with an increased error rate affecting operations such as creating new databases. This caused partial disruption for users attempting to provision databases on D1. The outage began at 4:20 PM UTC and was resolved by 5:10 PM UTC, lasting about 50 minutes. Cloudflare issued two official status updates during the incident, indicating investigation into D1 availability issues and later confirming the issue was resolved. The affected component was Cloudflare D1 (control plane).

Cloudflare carried out scheduled maintenance at the Newark edge location (EWR) on August 15, 2025. The maintenance window began at 5:00 AM UTC and lasted for about 5 hours, ending at 10:05 AM UTC. During the window, traffic routed through the Newark edge could experience intermittent issues, as indicated by the status page update that maintenance was in progress and updates would follow. The maintenance concluded with operations returning to normal by 10:05 AM UTC. The impact was confined to the Newark region with no broader outages reported.

Cloudflare carried out scheduled maintenance in the Prague (PRG) data center region on 2025-08-15. The maintenance window began around 02:30 UTC and ran for about an hour, with the status page indicating the work was in progress and potential impact. The Prague region transitioned to a degraded state at 02:35 UTC and then returned to normal at 03:35 UTC. The activity was regional to PRG and did not indicate a global outage. Cloudflare posted two status updates during the incident to inform customers that maintenance was underway and operations had resumed.

Cloudflare conducted scheduled maintenance in the Miami area (MIA). The maintenance window began around 6:00 PM UTC and the status transitioned to a maintenance state at 6:05 PM UTC. The window lasted approximately 4 hours and 5 minutes, with services returning to normal by 10:05 PM UTC. The incident was described as scheduled maintenance and was localized to the Miami region.

Cloudflare conducted a scheduled maintenance window in the Fukuoka region. The maintenance began at Aug 14, 2025 3:35 PM UTC and was completed by Aug 14, 2025 4:35 PM UTC. The provider published two status updates during the incident, including an in-progress note and a resolution indicating the region returned to normal operation. The work was regional in scope, affecting services in the Fukuoka area during the window and returning to normal after about one hour.