Cloud over-provisioning continues to be a major source of unnecessary spending for organizations. In this talk, Michael Gutierrez, Senior Manager of Cloud Platform at JSR, outlines how to detect and remediate resource inefficiencies using kube-prometheus-stack. Drawing from experience at JSR, Gutierrez explains why monitoring for over-provisioning is critical to achieving cloud cost efficiency, and how to build a targeted observability stack to support this goal.
The session highlights key metrics to track, how to configure alerts using Alertmanager for real-time insights, and how to visualize trends in resource usage with Grafana. It also explores how tools like Node Exporter and Prometheus can be used together to monitor CPU and memory consumption across clusters. Attendees will leave with actionable strategies for optimizing infrastructure and ensuring cloud environments remain both performant and cost-effective.