I spent a lot of time optimizing autoscaling on a GKE cluster at my previous company. During that process, I realized no tools were available to visualize how the Kubernetes scheduler perceives cluster nodes something that could help explain why a pod gets assigned to one node over another.
So, I created a simple Python application to visualize requested resources in a Kubernetes cluster. And that’s how k8sfoams was born.
It's a local, read-only dashboard that reads your ~/.kube/config, pulls node capacity and pod resource requests, and draws your cluster as a treemap: one box per node, one foam per pod, containers nested inside, and space where the free capacity actually is. No in-cluster deployment. No metrics-server.
One pip install.
pip install k8sfoams
It supports CPU or memory visualization on a 2D view.
Or both in 3D view.
It's rendered in pure CSS 3D, so no WebGL and no GPU required.
Check out the repository and let me know your thoughts:
Top comments (0)