Cost and Resource Optimization In Kubernetes
summary
Explains concrete rules and demos: if a cluster has 20 worker nodes but peak usage never exceeds six, scale back to save money; never scale a cluster down to one node for production and keep at least two nodes. Shows installation steps: run the installer, use kubectl port-forward and default admin/password for UI, and apply a Helm values.yaml. Compares tools with live demos — socivio self-managed (install and operate yourself) versus stormforge SaaS (portal, paid Optimize Pro) and shows cluster/namespace request vs optimized request metrics.
AI-written summary · about
chapters (9)
- 0:00intro
- 0:27preamble
- 0:38why?
- 3:24resource optimization
- 4:15scalability
- 5:06cost optimization
- 5:41various tools in the space
- 6:16demo
- 11:00thanks
related talks
15mDemo: Resource(Cost) optimization on planet scale Kubernetes clusters.
36mOptimizing Resource Usage in Kubernetes
11mK8s Resources Ensuring Performance and Optimization
28mScaling in Kubernetes Safely on On-Prem KaaS Across 1,300+ Clusters and 40,000+ Nodes
29mRight-Size Your Kubernetes Deployments
35mWhy Kubernetes Can't Get Around FinOps – Cost Management Best Pra...
44mFinOps at scale: Lessons from scaling Kubernetes workloads
59mHow Kubernetes optimisation can combat climate change