Cloudflare Pay Per Crawl
Cloudflare has confirmed that a major internal glitch caused the widespread internet outage on Tuesday. The company dismissed all rumours of a Distributed Denial of Service attack, which had quickly spread online after many top websites went down at the same time. Users across the world were unable to access platforms such as X, Canva and ChatGPT, and many received a Cloudflare error message. The sudden scale of the disruption made people believe that a major cyber attack was happening.
The company clarified that there was no malicious activity behind the incident. The problem began after an unexpected issue in one of its internal database systems triggered a chain reaction that affected how traffic was routed across its global network.
When the outage began, Cloudflare engineers initially thought it could be a large cyberattack because the early signs looked similar to heavy internet traffic spikes. Since many major websites stopped working at the same time, social media discussions also pointed towards the possibility of hackers being responsible.
After checking internal logs and systems, Cloudflare confirmed that there was no harmful activity. The real cause was a file that had grown larger than usual and had spread across the network. The team fixed the issue by stopping the file from spreading and restoring an older version.
The situation was different from the recent Microsoft Azure incident where the company blocked a very large attack on an Australian website. It was also different from the Amazon Web Services outage last month, which was caused by an issue in its traffic management system.
Also read: Cloudflare outage explained: Why services like Gemini, ChatGPT, Perplexity and Canva went down
Once engineers identified that the oversized file was responsible, they immediately stopped its spread and replaced it with an earlier version. By around 14:30 UTC, which was 07:00 p.m. IST, most core traffic had returned to normal.
The team continued to manage high traffic as services came back online. By 17:06 UTC, which was 10:36 p.m. IST, Cloudflare confirmed that all systems across its network were functioning normally.
Also read: Apple expands AppleCare+ in India for iPhone users: Check price, benefits and more
According to Cloudflare, the outage started on 18 November 2025 at 11:20 UTC, which was 04:50 p.m. IST. Matthew Prince, the co-founder of the company, explained the cause in a detailed blog post. He said the global outage took place after a change in database permissions created an unexpected flow of entries into a file used for its bot management system. The file became twice its normal size and was automatically shared across Cloudflare’s global servers.
The systems that read this file had a fixed size limit. Once the file crossed that limit, the software began to fail. This failure prevented the systems from routing traffic properly, which caused error pages for millions of users.
Also read: Best apps and games of 2025 on Google Play in India: District, CookieRun and more
Cloudflare co-founder Matthew Prince apologised for the disruption caused to customers and internet users worldwide. The company said that, because it manages a large part of the internet’s traffic, even brief periods of failure are unacceptable.
Cloudflare added that the incident was deeply disappointing and that steps are being taken to ensure that such a mistake does not happen again.