On Tuesday, a significant outage at Cloudflare disrupted access to several major websites, including X (formerly Twitter) and ChatGPT, for approximately three hours. The incident began around 6:20 a.m. ET when the company detected "unusual traffic" affecting its services. A status update was issued shortly after, indicating "internal service degradation."
Cloudflare's Chief Technology Officer, Dane Knecht, acknowledged the company's failure to its customers in a post on X, attributing the disruption to a software crash caused by an oversized configuration file intended for managing threat traffic. This file exceeded its expected size, which ultimately triggered the software malfunction.
Despite the outage's severity, Cloudflare stated that there was no evidence suggesting it was the result of a cyberattack or malicious activities. The company promised to provide a more detailed explanation of the incident through its blog in the near future.
Throughout the duration of the outage, affected sites experienced intermittent accessibility, with some users managing to reconnect briefly before losing access again. By approximately 11:45 a.m. ET, access to X, ChatGPT, and other previously impacted sites, such as Spotify, had been restored.
The outage coincided with scheduled maintenance at Cloudflare's data centers in several locations, including Atlanta and Los Angeles. This incident adds to a growing trend of internet outages, as reliance on a small number of content delivery networks has increased. An analyst from EMARKETER noted that these outages are becoming more frequent and prolonged, highlighting the strain on infrastructure due to rising demands from AI and streaming services.
As Cloudflare seeks to learn from this incident, it remains to be seen how it will address the challenges posed by increased internet traffic and the need for resilient infrastructure.