Skip to content

Namespace Management

Tenant admins can manage Kubernetes namespaces through the OICM+ admin tools. Namespaces let you isolate resources and workloads for different teams or projects.

Namespaces Management

1. List of Namespaces

The main page shows existing namespaces in a table format. Each entry includes:

  • Name – The namespace identifier.
  • Resource Allocation – CPU, memory, and GPU limits (if assigned).
  • Actions – Create, edit, or remove.

2. Creating a New Namespace

  1. Click Create Namespace.
  2. Enter a name.
  3. Define resource allocations (CPU, memory, GPU).
  4. Submit to finalize.

3. Accessing Kubernetes Configuration

After creating a namespace:

  1. Select the namespace.
  2. Click Generate Kubeconfig to download a namespace-specific config file.

Example: Use kubeconfig with kubectl: 

kubectl --kubeconfig=download_kubeconfig.yaml apply -f "manifest.yaml"

This config ensures all Kubernetes commands apply specifically to this namespace.

4. Important Considerations

  • Namespace Required Always reference the correct namespace in your Kubernetes manifests.
  • Resource Isolation CPU, memory, and GPU settings help prevent resource conflicts among tenants.

5. Example: Deploying JupyterLab

apiVersion: apps/v1
kind: Deployment
metadata:
  name: jupyterlab
  namespace: <YOUR_NAMESPACE_NAME>
labels:
    app: jupyterlab
spec:
  replicas: 1
  selector:
    matchLabels:
      app: jupyterlab
  template:
    metadata:
      labels:
        app: jupyterlab
    spec:
      containers:
      - name: jupyterlab
        image: cschranz/gpu-jupyter:v1.6_cuda-11.8_ubuntu-22.04_python-only
        ports:
        - containerPort: 8888
        env:
        - name: JUPYTER_ENABLE_LAB
          value: "yes"
        - name: GRANT_SUDO
          value: "yes"
  • Save as jupyterlab_deployment.yaml.
  • Apply:
kubectl --kubeconfig=download_kubeconfig.yaml apply -f jupyterlab_deployment.yaml

This spins up a JupyterLab instance within your specified namespace.

Next Steps