Submit Kubernetes Resources

Kubeflow assigns the default-editor service account to the Jupyter notebook Pods. This service account is bound to the kubeflow-edit ClusterRole, which has namespace-scoped permissions to the many Kubernetes resources including:

• Pods
• Deployments
• Services
• Jobs
• TFJobs
• PyTorchJobs

You can get a full list of these permissions using:

kubectl describe clusterrole kubeflow-edit

You can therefore create the above Kubernetes resources directly from your Jupyter notebook in Kubeflow. The Kubernetes kubectl command-line tool is pre-installed in the notebook.

Run the following command in a Jupyter notebook cell to create Kubernetes resources:

!kubectl create -f myspec.yaml

The myspec.yaml file should describe one of the above Kubernetes resources. For information about the format of the YAML file, see the Kubernetes object guide.

Next steps

• See the guide to setting up your Jupyter notebooks in Kubeflow.
• Explore the components of Kubeflow, including custom Kubernetes resources.