Cloudflare Down: When a “Hidden” Internet Giant Stumbles, Half the Web Falls Over
Cloudflare Down: If you tried to load X (formerly Twitter), ChatGPT, Spotify, League of Legends, or countless other sites this morning and were greeted with cryptic Cloudflare errors (“Please unblock challenges.cloudflare.com” or plain 500 Internal Server Errors), you weren’t alone.
At approximately 11:20 GMT (6:20 a.m. ET), Cloudflare — one of the largest and most critical pieces of invisible internet plumbing — suffered a widespread incident that cascaded across millions of websites and apps.
What actually happened? Cloudflare Down
Cloudflare’s own status page and public statements describe it as:
- A sudden “spike in unusual traffic” to one of its internal services
- This spike caused elevated error rates (mostly HTTP 500-class errors) across multiple Cloudflare services globally
- While the majority of traffic continued to flow normally, a significant portion of requests failed
- Services began recovering within roughly an hour, but residual elevated errors persisted for some time
By 12:30 GMT the company reported: “We are seeing services recover, but customers may continue to observe higher-than-normal error rates as we continue remediation efforts.”
As of 13:30 GMT, Cloudflare says the immediate issue has been mitigated and they are now shifting focus to root-cause analysis.
Who was affected?
Downdetector lit up like a Christmas tree for:
- X (Twitter)
- OpenAI / ChatGPT
- Spotify
- Discord
- League of Legends & other Riot Games titles
- Shopify stores
- Crunchyroll
- Many news sites, crypto exchanges, and SaaS dashboards
Even Downdetector itself was intermittently unreachable because… it sits behind Cloudflare.
Why one company’s hiccup can break half the internet
Cloudflare is often called “the biggest company you’ve never heard of.” It:
- Operates one of the world’s largest anycast networks (270+ cities, 330+ Tbps capacity)
- Terminates ~20–25 % of all global HTTP/HTTPS traffic on an average day
- Provides DDoS protection, CDN, DNS, Zero Trust, and bot management, and more
- Serves everyone from tiny blogs to FAANG-scale giants
When something goes wrong at that layer, the blast radius is enormous — exactly what we saw today.
Professor Alan Woodward (Surrey Centre for Cyber Security) summed it up neatly to The Guardian: “We’re seeing how few of these companies there are in the infrastructure of the internet, so that when one of them fails it becomes really obvious quickly.”
Was it a cyber-attack?
Cloudflare says it does not believe so at this stage, and experts agree it’s unlikely. A company of Cloudflare’s scale and paranoia is engineered with no single point of failure for DDoS or similar attacks. The incident appears internal — possibly related to the unusual traffic spike they mentioned, or coincidentally overlapping with scheduled maintenance in several datacenters (Tahiti, Los Angeles, Atlanta, Santiago).
This isn’t the first time — and it won’t be the last
- June 2024: Cloudflare configuration error took down Google, Amazon, Spotify, etc. for ~40 minutes
- November 2024: AWS us-east-1 partition outage brought down thousands of services
- July 2022: Cloudflare BGP leak caused 90-minute global degradation
The pattern is clear: the modern internet infrastructure is incredibly resilient 99.999 % of the time, but when one of the handful of hyperscale backbone providers stumbles, the effects are immediate and spectacular.
Takeaway for the Rest of Us
- Diversify your critical dependencies when possible (multi-CDN, multi-cloud DNS, etc.)
- Have offline/fallback experiences ready (many mobile apps completely failed today because they couldn’t reach Cloudflare-protected APIs)
- Remember that “the cloud” is just someone else’s computer — and sometimes that computer has a bad day.
Cloudflare has an excellent track record of transparent post-mortems. We’ll likely get a detailed technical write-up within the next few days explaining exactly what caused today’s spike in “unusual traffic.”
Until then, the internet is mostly back — but it was a stark reminder of how much of the web we all love runs on surprisingly few foundation stones.
