By Half of the Internet just went down today, November 18th 2025, as Cloudflare was hit by a major network issues inside their infrastructure.
Outages resolved
As of writing this article – for which we needed to disable Cloudflare's "Proxy" service – there are still global outages reported. – Update 13:20 UTC
We see recoveries across a lot of sites now – Update 14:35 UTC
It has been stable for websites using Cloudflare in the past 20 minutes. The Cloudflare Status Page also confirms that a fix has been rolled out – Update 14:50 UTC
The Cloudflare Dashboard still seems to have issues, for example logging in. – Update 15:25 UTC
Dashboard seems to work again without errors – Update 16:00 UTC
Widely used CDN – global impact for everyone involved
Cloudflare is a global CDN provider, one of the largest of the whole Internet. An expected 20-25% of websites use Cloudflare as CDN provider. At Geeker's Digest we also use the services of Cloudflare and our website was down for quite some time (we had to disable Cloudflare to get our site back online).
Note: Cloudflare Free Tier accounts were definitely affected by the outages. At least some, maybe all, Enterprise accounts were not affected by the outage.
A website visitor should see an error 500 in the browser. The error correctly identifying a problem inside Cloudflare's infrastructure services.
But also major services and websites are affected. This also includes X (formerly Twitter) as it stopped working entirely or had issues with the feed or posting new content.
Incident Status Updates
The Cloudflare Status Page confirmed the global network issue:
As of writing this article, our website again went flapping (up -> down -> up -> down) and there are still issues reported, including the Cloudflare Dashboard and API.
As of 12:40 UTC there were still global issues reported.
Cloudflare updated the incident information, which is currently being investigated.
Most services have recovered – but not all
Most of the Cloudflare services – and therefore also the sites using Cloudflare – have recovered at around 14:30 UTC. However some problems or after-effects are still happening. This also includes the Cloudflare Dashboard. For example login is not correctly working or authentication (API) errors to access domains/zones could show up.
This error was just seen a few minutes ago, at 15:25 UTC.
The Cloudflare Status page confirms there are still issues with the Dashboard.
At 16:00 UTC we see no more issues inside the Cloudflare Dashboard. We have also enabled Cloudflare Proxy again for our own website.
Post mortem analysis
In the meantime the CTO of Cloudflare, Dane Knecht, has shared the reason for the outage on LinkedIn:
In short, a latent bug in a service underpinning our bot mitigation capability started to crash after a routine configuration change we made. That cascaded into a broad degradation to our network and other services. This was not an attack.
Some public posts on social media or on news articles mentioned a traffic spike as the reason for the outage – which was completely wrong.
Cloudflare is known for their very transparent and open communications. There will likely be a full post-mortem of this incident on their company blog in the coming days.
Update November 19th:
As expected, a full post-mortem article about the incident has been posted. It explains in technical detail what went wrong.
Kudos to Cloudflare for doing this. There's rarely a company being so transparent about internal processes (and mistakes).
