Cloudflare outage caused by bad software deploy (updated)

What this post added

This post details a critical incident where a misconfigured regular expression in a new Cloudflare WAF Managed rule caused a global outage due to 100% CPU utilization across the network. It describes the simulated mode deployment, the impact on traffic (82% drop), the rapid rollback ('global termination' of WAF Managed Rulesets), and the subsequent review and re-enabling of the rules. The post emphasizes the insufficiency of existing testing processes and the commitment to improving them to prevent future incidents.