Explore a critical incident analysis in this SREcon23 Europe/Middle East/Africa conference talk where a single line change in a configuration management system led to the unavailability of thousands of websites. Delve into the infrastructure that mitigated widespread damage, extract valuable lessons on infrastructure design and operational procedures, and discover significant improvements implemented in the aftermath. Gain insights into Kubernetes infrastructure, focusing on operators, automation, manual intervention, configuration management, and backup strategies. Learn from CERN experts Francisco Borges Aurindo Barros and Jack Henschel as they dissect the intense recovery procedure and share their experiences in managing large-scale web infrastructures.