K8s Workload Autoscaling Performance Dimensions

Explore the complexities of applying Cloud Cost Optimization, Savings & FinOps practices to production Kubernetes environments in this 26-minute conference talk from DevConf.CZ 2024. Delve into the main factors causing waste, overcommitment, and resource scarcity across CPU, Memory, and Network Bandwidth. Learn from real-world production issues and their solutions, gaining insights into the necessary optimizations for clusters and applications. Understand why blindly trusting various autoscalers and Kubernetes schedulers out of the box may not be sufficient, and discover strategies for balancing cost optimization with performance and reliability in your Kubernetes workloads.