Large-scale Internet outage: popular platforms are down due to problems with Cloudflare
18 November 2025 14:35
Today, users in Ukraine and around the world have faced massive outages on popular Internet platforms.
The instability was caused by technical problems at Cloudflare, one of the world’s largest providers of cloud security and content delivery networks (CDNs), reports "Komersant Ukrainian".
What’s happening now
Monitoring services and users report that , there are significant connection delays, 500 errors or complete unavailability of the following resources:
- X (Twitter): news feed and media files are not loading.
- Spotify: problems playing music and logging into the app.
- ChatGPT: OpenAI chatbot does not respond to requests or generates network errors.
- Discord: delays in messages and connection problems.
- Downdetector: ironically, the site that monitors failures is itself experiencing interruptions due to the influx of users and dependence on Cloudflare infrastructure.
Ukrainian news sites and online stores that use Cloudflare’s DDoS protection may also be affected.
Similar large-scale failures have already occurred before:
October 2023: A short-term but powerful outage that affected Discord, Omegle, and NordVPN services. At the time, the company attributed it to problems with internal configuration management tools.
June 2022: One of the largest outages. Due to a software update error, hundreds of popular services, including FTX, Discord, Shopify, and the Cloudflare website, went down in 19 Cloudflare data centers. The problem was localized in 30 minutes, but it took several hours to restore service.
July 2020: The global blackout was caused by a router configuration error in the company’s backbone network. This led to a 50% drop in traffic worldwide.
What should users do?
Cloudflare engineers are currently working to fix the problem.
Users are advised to:
- Do not refresh pages too often (this creates an additional load).
- Use a VPN (sometimes changing the region helps to bypass the problematic node, although in the case of Cloudflare it is not always effective).
- Just wait – usually such failures are resolved within 30-60 minutes.
