Alauda AI

Create WorkspaceKind

TOC

Prerequisites

  • Ensure you are Cluster Admin.
  • Ensure you have kubectl configured and connected to your cluster.
  • Ensure you have created SA.

Procedure

Create SA

Create a file workbench-sa-with-rolebinding.yaml with following lines, and run kubectl apply -f workbench-sa-with-rolebinding.yaml to create service account used by workbench instances in the namespace.

DETAILS
apiVersion: v1
kind: ServiceAccount
metadata:
  name: workbench-editor
  namespace: NAMESPACE_NAME
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: workbench-sa-binding
subjects:
- kind: ServiceAccount
  name: workbench-editor
  namespace: NAMESPACE_NAME
roleRef:
  kind: ClusterRole
  name: aml-namespace-editor
  apiGroup: rbac.authorization.k8s.io
  1. NAMESPACE_NAME should be same with the namespace of workbench created.
  2. Same as above.

Create WorkspaceKind using CLI

WARNING
  • When use jupyterlab_v1beta1_workspacekind.yaml, please change CLUSTER_NAME in NB_FREIX to your target cluster name, which is WorkspaceKind will be created at.
  • When creating WorkspaceKind, image URL must be changed to the actually one for the current local cluster !!!