Today wasn’t normal for the internet.
One moment everything was fine.
Next moment, half the apps we use daily stopped working.
ChatGPT wasn’t loading.
X (Twitter) threw errors.
Udemy, Canva, all struggling.
If your apps felt broken, the problem wasn’t with your WiFi.
It was Cloudflare.
And when Cloudflare goes down, the internet feels it everywhere.
I checked updates from Reuters, major news outlets, and Cloudflare’s own status page.
Here’s the simplest developer-friendly summary of what happened today.
Cloudflare sits between users and most websites
When you open a website, your request usually goes to Cloudflare first.
Companies use it for:
- CDN
- WAF
- DDoS protection
- Caching
- Routing
- API security
So when Cloudflare slows down, every app behind it slows down too.
What actually went wrong today
Cloudflare shared that one of their internal services received unexpected traffic.
Not from attackers.
Not from unknown IPs.
But from their own network, just at a scale they didn’t expect.
They didn’t name the service, but it was critical enough that once it got overloaded, other systems started failing.
That’s why:
- ChatGPT
- X
- Udemy
- Canva
all showed the same error:
“We can’t reach the server right now.”
It wasn’t their fault.
The layer beneath them was unhealthy.
This outage hit developers too
A lot of CI/CD pipelines broke because dependency downloads failed.
Examples:
- Java JAR downloads
- npm installs
- Docker image pulls
- API calls through Cloudflare
Even if your app doesn’t use Cloudflare directly, some service your tools depend on probably does.
This is how interconnected the internet has become.
Cloudflare pushed a fix quickly
Around 14:42 UTC, Cloudflare rolled out a fix.
By 14:57 UTC, most services started coming back.
Some people still saw small glitches afterward, but the worst was over quickly.
Cloudflare will publish a full RCA soon, which will explain exactly what happened.
I’ll write a follow-up post when that drops.
For developers, there’s a lesson in this
Incidents like today remind us:
- One dependency can break many systems
- Monitoring should cover underlying infrastructure, not just your app
- Fallback paths matter
- External CI/CD dependencies can fail anytime
The modern internet rests on a few critical layers and Cloudflare is one of the biggest.
What was your experience?
Did your app break?
Did your build pipeline fail?
Did your users complain?
Drop your experience below it’ll help others understand the impact.
If you want calm, simple breakdowns of incidents like this, or stories about cloud, dev, and building real-world systems — you can follow me here.
Top comments (0)