Cloudflare is working to resolve a major outage that briefly knocked out access to parts of the global internet, disrupting platforms such as X, Facebook, Spotify, Canva, and other popular services. The company confirmed it was dealing with a significant network problem affecting multiple customers, with the first alert posted at 11:48 UTC. Cloudflare explained that its support portal was also experiencing errors due to problems with the provider behind it, which meant some users could not view or respond to support cases. However, customer inquiries were still being handled through live chat and the emergency phone line.

Cloudflare said it was in contact with the third party involved to fully assess the disruption and limit its impact. As one of the largest internet infrastructure networks, Cloudflare handles traffic for millions of websites and processes around 81 million HTTP requests per second. Users trying to access sites reliant on the service were often met with internal server error notifications, urging them to retry later. The outage also affected Down Detector, which nonetheless showed a sharp rise in reports as soon as it loaded properly.

According to early indications, the issue generated widespread 500 errors on websites, and Cloudflare noted that both its dashboard and API had been failing. At the same time, many users reported that platforms such as X, Spotify, Letterboxd, and Grindr were becoming inaccessible. The situation unfolded just hours after Cloudflare announced scheduled maintenance in a Santiago datacenter, though it remains unclear whether that work contributed to the disruptions.

By 12:21 UTC, Cloudflare reported that services were beginning to recover, although higher error rates could continue while the team worked through the remaining problems. The event followed recent large scale outages at Amazon Web Services, which had knocked out apps including Snapchat, Reddit, Fortnite, Canva, and Duolingo. During the AWS problems, many users were unable to log into accounts or access basic features until Amazon eventually traced the fault to a core part of its North Virginia cloud operations.

This latest disruption began around 11:20 am, with Downdetector registering more than 11,000 affected users within about twenty minutes, highlighting how quickly instability in a major network provider can ripple across the online ecosystem.