Rapidly Scaling for Breaking News with Karpenter and KEDA

Explore how The New York Times handles rapid scaling for breaking news events in this 47-minute conference talk. Learn about the challenges of doubling Kubernetes cluster nodes in under two minutes to manage traffic spikes reaching over 100,000 requests per second. Discover the implementation of Karpenter and KEDA in multi-tenant EKS clusters, focusing on scaling solutions for the unified HTTP ingress bottleneck. Gain insights into design decisions balancing flexibility, maintainability, and cost-effectiveness in a high-pressure news environment. Understand the intricacies of scaling infrastructure to meet sudden, short-lived traffic surges triggered by breaking news alerts and popular content like crossword puzzles.