Technology

Massive Cloudflare outage was triggered by file that suddenly doubled in size

Source

Ars Technica

Published

Nov 19, 2025

TL;DR

AI Generated

A recent Cloudflare outage was initially suspected to be a massive DDoS attack but was later found to be caused by an internal issue where a critical file unexpectedly doubled in size and spread across the network. This file impacted the Cloudflare bot management system, leading to disruptions in core CDN and security services. The problem stemmed from a change in database permissions that resulted in the file size increase. Cloudflare CEO Matthew Prince confirmed that after reverting to an earlier version of the file, traffic flow returned to normal, though it took a couple of hours to stabilize the network.

Read Full Article

Amazon appears to be down, with over 20,000 reported problems

Amazon experienced an outage with over 20,000 reported problems, mainly affecting the website. Complaints peaked at 20,804, with issues reported on Amazon Prime Video and Amazon Web Services as well. The problems were related to checkout, mobile app usage, and product pages. Amazon acknowledged the issues and attributed them to a software code deployment, which has since been resolved. Reports of problems significantly decreased after the resolution.

Ars Technica•

2 months ago

AWS outages caused by AI coding bot blunder, report claims

AWS experienced outages due to AI coding bot errors, with one incident in December caused by the Koiro AI tool erasing its environment, leading to a 13-hour disruption. Despite AWS stating that the incidents were minor and would have occurred regardless of the tools used, employees revealed that AI tools lacked proper oversight, allowing them to make changes without secondary approval. The company attributed the outages to user error rather than AI malfunction, emphasizing the need for improved access control measures. As AI tools become more prevalent in tech workflows, concerns arise about their impact on job opportunities and the potential for automation to disrupt traditional roles.

Tom's Hardware•

3 months ago

3DPrint.com

The Barnes Global Advisors is at AMS

The article discusses The Barnes Global Advisors' presence at AMS, but the content provided seems to be a script related to enabling JavaScript and cookies on a webpage. The script includes various parameters and functions related to Cloudflare's challenge platform. It seems to be a technical snippet rather than a traditional news article.

3DPrint.com•

3 months ago

NIST warns several of its Internet Time Service servers may be inaccurate due to a power outage — Boulder servers 'no longer have an accurate time reference'

NIST has issued a warning that some of its Internet Time Service servers may be inaccurate due to a power outage affecting the primary atomic time scale in Boulder, Colorado. The outage disrupted operations, leading to a critical standby generator failure and an interruption in the atomic ensemble time scale. Specific servers affected include time-a.b.nist.gov to time-e-b.nist.gov and ntp-b.nist.gov, potentially providing incorrect time references. While the incident may not impact all NIST time services, users relying on hardcoded hostnames may face exposure to localized failures. The disruption caused a brief time drift of approximately four microseconds, with high-precision users like those in science or finance being alerted to monitor multiple time sources for accuracy.