Cloudflare Has Finally Explained What Happened in the Massive Outage

Summary is AI-generated, newsroom-reviewed

New Delhi: Cloudflare has finally explained what caused the massive outage that disrupted the internet on November 17 and 18. The company confirmed that the issue was not a cyber attack or any kind of malicious activity. Instead, it was a technical problem inside its own systems.

The outage affected many popular platforms worldwide. X (formerly Twitter), Spotify, ChatGPT, Perplexity, Gemini, Canva, Letterboxd, Bet365, League of Legends, Sage, and Downdetector all went down. Users reported timelines failing to load, posts not going through, and apps refusing to connect. For hours, millions of people struggled to access their favorite services.

Cloudflare explained that the problem started with a change to one of its database systems. This change caused the database to output multiple entries into a “feature file” used by its Bot Management system. The file doubled in size unexpectedly. That oversized file was then sent across all machines in Cloudflare’s global network.

The software running on these machines reads the feature file to keep track of online threats. But the software had a limit on how large the file could be. When the file exceeded that limit, the software failed. This failure disrupted the routing of internet traffic across Cloudflare’s network.

At first, engineers suspected a massive DDoS attack. Later, they realized the real cause was the oversized feature file. Cloudflare stopped the spread of the file and replaced it with an earlier version. By 2:30 pm on November 18, core traffic was flowing normally again. By 5:06 pm, all systems were back to normal.

Cloudflare apologized for the disruption. The company said it knows how important its role is in the internet ecosystem. It admitted that any downtime is unacceptable and called the outage “deeply painful.”

For users, the incident showed how dependent modern apps and websites are on Cloudflare’s infrastructure. A single technical glitch at the company can ripple across the internet, taking down multiple services at once.

On November 17, the outage first hit several platforms. X (formerly Twitter) briefly came back online but went down again for many desktop users at 6:27 pm, with timelines failing to load and posts not going through. Engineers were still investigating at the time, and the incident was described as a developing story. Alongside X, services like Spotify, ChatGPT, Perplexity, Gemini, Canva, Letterboxd, Bet365, League of Legends, Sage, and Downdetector also faced disruptions.

This sequence of events shows how the outage unfolded over two days, with widespread impact before Cloudflare identified and fixed the root cause.