Demystifying Kubernetes resource management

It's a Wrap!

Thanks for joining us at AWS on January 30th.  You can relive the event by accessing presentation materials here:

John Platt's slides on Slideshare  || Rodrigo Bersa's resources on Github

Event Summary

We'll explore Kubernetes resource management, including its basics, best practices, and real-world situations. We'll discuss how requests and limits simplify Linux kernel capabilities and affect Kubernetes scheduling. We'll cover Quality of Service (QoS) classes, explaining how Kubernetes measures usage and handles node pressure. Finally, we'll demonstrate how to use Resource Quotas and Limit Ranges to manage large-scale Kubernetes deployments. We'll show how Karpenter can be used to simplify management by launching the right compute resources for the application workloads requests while responding quickly and automatically to changes in application load and resource requirements. Karpenter is an open-source project which observes the aggregate resource requests of unscheduled pods and makes decisions to launch and terminate nodes to minimize scheduling latencies and infrastructure cost.