Cloudflare Outage Disrupts Global Websites, Caused by Internal Database Error

Cloudflare experienced its most severe outage since 2019 on Tuesday, temporarily disrupting access to millions of websites globally. CEO and co-founder Matthew Prince described the incident as the worst in over six years, affecting users across sectors, including news media and e-commerce platforms in Nigeria. The company confirmed that core traffic was largely restored by mid-afternoon.

The disruption occurred when an internal change to a database connected to Cloudflare’s Bot Management system caused a critical error. The system generated a larger-than-expected feature file, which overwhelmed the software responsible for routing traffic. “The issue was not caused by a cyberattack or malicious activity,” Cloudflare emphasized, clarifying widespread concerns about security breaches.

Several global platforms, including X, ChatGPT, and Downdetector, experienced connectivity issues during the outage. Businesses relying on Cloudflare’s bot rules inadvertently blocked legitimate traffic, while sites not using these rules continued to operate. Analysts compared the outage to recent disruptions at Microsoft Azure and Amazon Web Services, highlighting the risks of relying heavily on a small number of internet infrastructure providers.

In response, Cloudflare outlined immediate measures to prevent similar incidents, including improving the handling of configuration files, adding global feature kill switches, and reviewing failure modes across core proxy modules. These steps aim to strengthen resilience as internet services become increasingly centralized around key providers.

Prince acknowledged that as global internet traffic continues to consolidate, incidents of this scale may be difficult to fully prevent. Still, he reassured users that Cloudflare remains committed to reliability and minimizing disruptions for the millions of websites that depend on its network every day.
source: Nairametrics