Node Monitoring

The node list on the dashboard shows per-node resource usage in real time, powered by the Kubernetes metrics-server.

Node Metrics

Each node entry in the list displays:

Field	Description
Name	Kubernetes node name
Status	Ready or NotReady
CPU	<used> / <capacity> cores with a blue progress bar
Memory	<used> / <capacity> GB with a violet progress bar

The progress bars fill proportionally to usage. A bar near 100% indicates a heavily loaded node.

Enabling Metrics

Node metrics require metrics-server to be installed in the cluster. If metrics-server is not available, the progress bars are hidden and only capacity figures are shown.

Installation

# Standard installation
kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml

# For k3d or local clusters (disable TLS verification):
kubectl patch deployment metrics-server -n kube-system \
  --type=json \
  -p='[{"op":"add","path":"/spec/template/spec/containers/0/args/-","value":"--kubelet-insecure-tls"}]'

Verify

kubectl top nodes

If this command returns node metrics, the dashboard will automatically show real-time data.

Workload-Level Metrics

In addition to node-level metrics, the Jupyter Notebooks and MCP Servers pages show a Usage column with per-workload CPU (millicores) and memory (MB) consumption.

These are aggregated from all pods belonging to each workload deployment.