Explore a real-world case study of a metastable failure in a globally distributed CDN through this 25-minute conference talk from SREcon23 Americas. Delve into the challenges faced during an extended system outage, understand the design changes implemented to prevent future incidents, and discover the unconventional solution that ultimately led to recovery. Learn valuable insights applicable to addressing metastable failures in various systems, gaining practical knowledge for improving incident response and system resilience in large-scale distributed environments.
Overview
Syllabus
SREcon23 Americas - We're Still Down: A Metastable Failure Tale
Taught by
USENIX