What caused the Cloudflare outage?
The outage was caused by a bug in Cloudflare's Bot Management system, triggered by a database permissions change that led to an oversized feature file.
Yanuki
ARTICLE DETAIL
Cloudflare Outage Disrupts Internet Services: What Happened and Why It Matters | Discord Outage Disrupts Services; Accenture Acquires Ookla to Enhance Network Intelligence | TCS Launches Gemini Experience Center in US with Google Cloud | Tencent QClaw Enables Dual Access to WeChat and QQ | OpenClaw AI Agents Surge in Popularity Amidst Security Concerns | Apple at 50: The Untold Story of the iPhone | Privacy Concerns Rise Over Meta's AI Smart Glasses | Apple Unveils MacBook Air with M5 Chip and Renames CPU Cores | TikTok Outage: Impact, Causes, and How Brands Can Prepare | Cloudflare Outage Disrupts Internet Services: What Happened and Why It Matters | Discord Outage Disrupts Services; Accenture Acquires Ookla to Enhance Network Intelligence | TCS Launches Gemini Experience Center in US with Google Cloud | Tencent QClaw Enables Dual Access to WeChat and QQ | OpenClaw AI Agents Surge in Popularity Amidst Security Concerns | Apple at 50: The Untold Story of the iPhone | Privacy Concerns Rise Over Meta's AI Smart Glasses | Apple Unveils MacBook Air with M5 Chip and Renames CPU Cores | TikTok Outage: Impact, Causes, and How Brands Can Prepare
Technology / Internet Outages
Cloudflare Outage Disrupts Internet Services: What Happened and Why It Matters
On November 18, 2025, a significant outage at Cloudflare, a major internet infrastructure provider, caused widespread disruptions across numerous online services and websites. This incident highlights the critical role Cloudflare plays in t...
Key Insights
- **Widespread Impact:** The outage affected major platforms including X (formerly Twitter), Spotify, OpenAI, and Amazon Web Services.
- **Root Cause:** A bug in Cloudflare's Bot Management system, triggered by a database permissions change, led to the generation of an oversized feature file that overwhelmed the network.
- **Rapid Response:** Cloudflare's team identified and resolved the issue within a few hours, restoring core traffic flow by 14:30 UTC and fully resolving all systems by 17:06 UTC.
- **Vulnerability of Centralized Infrastructure:** The incident underscores the fragility of relying on a small number of providers for critical internet services. Why does this matter? Outages like these can disrupt business operations, hinder communication, and impact user experiences across the globe.
In-Depth Analysis
### Background Cloudflare acts as a critical intermediary for a significant portion of internet traffic, providing CDN, security, and DNS services. Its outage cascaded through the internet, demonstrating how interconnected the web has become.
### Technical Breakdown The outage was triggered by a change in a ClickHouse database system's permissions, causing the Bot Management system to generate an unexpectedly large "feature file." This file, propagated across the network, exceeded software limits, leading to system failures. The issue manifested as HTTP 5xx errors and increased latency for users.
### Impacted Services - **Core CDN and Security Services:** HTTP 5xx status codes were widely served. - **Turnstile:** Failed to load. - **Workers KV:** Returned elevated HTTP 5xx errors. - **Dashboard:** Most users were unable to log in. - **Email Security:** Temporary loss of access to an IP reputation source. - **Access:** Widespread authentication failures.
### Remediation Steps Cloudflare addressed the issue by: 1. Stopping the generation and propagation of the bad feature file. 2. Manually inserting a known good file into the distribution queue. 3. Restarting the core proxy.
### Lessons Learned Cloudflare is implementing measures to prevent future incidents, including: - Hardening ingestion of Cloudflare-generated configuration files. - Enabling more global kill switches for features. - Eliminating the ability for core dumps to overwhelm system resources. - Reviewing failure modes for error conditions across all core proxy modules.
### How to Prepare - **Diversify Services:** Don't rely on a single provider for all critical services. - **Monitor Status Pages:** Keep an eye on the status pages of your key service providers. - **Implement Redundancy:** Where possible, implement redundant systems to ensure business continuity.
### Who This Affects Most Businesses and individuals who rely heavily on Cloudflare for their online presence and services were most affected by this outage.
FAQ
How long did the outage last?
Core traffic was largely restored by 14:30 UTC, with all systems functioning normally by 17:06 UTC on November 18, 2025.
What services were affected?
Services affected included core CDN, Turnstile, Workers KV, the Cloudflare Dashboard, Email Security, and Access.
Takeaways
- Cloudflare's outage highlights the internet's reliance on a few key infrastructure providers.
- A seemingly small configuration change can have widespread consequences.
- Monitoring service status and diversifying providers can help mitigate the impact of future outages.
Discussion
Do you think incidents like this will become more or less frequent in the future? Let us know in the comments!
Share this article with others who need to stay ahead of this trend!
Sources
Disclaimer
This article was compiled by Yanuki using publicly available data and trending information. The content may summarize or reference third-party sources that have not been independently verified. While we aim to provide timely and accurate insights, the information presented may be incomplete or outdated.
All content is provided for general informational purposes only and does not constitute financial, legal, or professional advice. Yanuki makes no representations or warranties regarding the reliability or completeness of the information.
This article may include links to external sources for further context. These links are provided for convenience only and do not imply endorsement.
Always do your own research (DYOR) before making any decisions based on the information presented.