Join us as we discuss a strategic method of autonomously and continuously optimizing your K8s environment with a combination of Karpenter, Keda, and pod right-sizing\n\nHPA, VPA, and Cluster autoscalers help us efficiently manage our clusters in a way that ensures resilience and availability using as few resources as possible. However, this is easier said than done due to a critical problem that many organizations are facing - Pod right-sizing.\n\nThe challenge with continuous, and effective pod right-sizing stems from multiple factors. Shifting left resource allocations, seasonality, and continuous delivery are introducing constant changes into the environment. To keep your environment continuously optimized for both cost and performance, it takes a strategic approach, mixed with the right tooling to drive optimal results.\n\nIn this session, we will discuss how the combination of Karpenter, Keda, and PerfectScale can safely and effectively right-size your pods, horizontally scale your workloads, and optimize your nodes automatically, keeping your app performance high and your cloud bills low.

This session will provide a comprehensive overview of the node autoscaling technologies in the Azure Kubernetes Service. Philip will cover everything you need to know about the default node scaler in AKS, cluster autoscaler and the new sheriff in town: Karpenter. We will furthermore explore the differences between these technologies in a live demo and take a look at the pros and cons.