Notebooks

Prev Next

Available in VPC

Get an overview the Notebooks interface. From Notebooks, you can create, delete, or access running notebooks to monitor the learning process. You can also view the details and recent event history of running notebooks.

Notebooks

The Notebooks interface includes the following components:

mlxp_console_notebooks01_ko

Component Description
① Menu name Current menu name.
② Basic features Add/Delete a notebook.
③ Notebook list View the list and details of running Notebooks.

View notebook details

The list of notebooks created by the user includes:

  • Notebook name: Name of the notebook set when initially created.
  • Date of creation: Date of initial creation.
  • Status: Status of the notebook.
  • Type: Application selected to run the notebook at creation.
    • JuPyterLab
    • VisualStudio Code
  • Image: Image selected when creating the notebook.
  • Actions: Access the created notebook or change its status.
    • CONNECT: Click to access the notebook runtime environment in a new window.
    • STOP: Disable the notebook.
    • Start: Enable the notebook.

Create a notebook

To create a new notebook:

  1. Click [Add].
  2. When the Create notebook page appears, enter a name for the notebook.
    • The name should be between 3 and 30 characters long and include lowercase letters (a-z), numbers (0-9), and hyphens (-).
    • Begin with a letter and end with a letter or number.
    • The livy keyword cannot be used in the name.
    • Duplication is not allowed.
  3. Select an application to run the notebook.
    • JuPyterLab
    • VisualStudio Code
  4. Select an image. Images available on the ML expert platform include:
    • kubeflow-jupyter-*: Jupyter-lab based on official Kubeflow images.
    • kubeflow-codeserver-*: Codeserver based on official Kubeflow images.
  5. Set the CPU and memory specifications to be used for the notebook.
Note
  • The minimum values correspond to Kubernetes resource requests, while the maximum values correspond to Kubernetes resource limits.
  • As scheduling issues may occur if the minimum values are set too high, it is recommended to use the default values provided.
  1. Set the GPU instance specifications to be used for the notebook.
  2. Set up a workspace volume. For your workspace volume, you can create a new volume or use an existing volume.
Caution
  • Only a single volume can be mounted.
  • When setting up a workspace volume, the mount path may not be the same as the user’s home path within the Notebook image.
  • If you set the workspace volume mount path to the image’s home path (for example, /home/irteam), it can overwrite existing data located at that home path within the image. If you need to use profiles, utilities, or other data stored in the image’s HOME PATH, you should set the workspace volume mount path differently from the image’s home path.
  • You can run a notebook without mounting a workspace volume. However, since the internal storage of the container is used, any data is lost when the notebook is terminated.
  1. Set up a data volume. Mount a separate data volume on the notebook and use it to import data or save your activity history. If you do not need a separate data volume, you can skip this step. For your data volume, you can create a new PVC or use an existing PVC.
Note
  • Multiple volumes can be mounted.
  • To keep your work results produced in a notebook on separate storage, you can create and mount a data volume.
  1. Select Configurations from Advanced Settings, if necessary.
Caution
  • Since the ML expert platform currently provides GPU resources to private zones, select the GPU zone information assigned to your workspace.
  • For GPU zone information, see View available GPU zone information.
  1. Configure the Miscellaneous settings option in Advanced Settings, if necessary.
  2. Click [Launch]
Caution
  • A local path volume provides NVMe storage. NVMe storage is physically connected to a specific host and is assigned directly to a GPU instance.
  • When a GPU instance is moved to another host via failover or host migration, the NVMe storage attached to the original host does not automatically migrate. As a result, you cannot access the existing NVMe data from the GPU instance that is moved to a different host.
  • We recommend using Data Manager, NCloud Storage, or Object Storage for data that require long-term retention.

Delete a notebook

To delete a running notebook:

  1. Select the notebook you want to delete and click [Delete]
  2. In the Delete notebookpopup window that appears, enter the name of the notebook to delete and click [Delete]
  3. Verify that the selected notebook has been deleted from the Notebook list.

View notebook details

You can view the details of the selected notebook. The details are divided into tabs.

Overview

These tabs include:

  • Volumes: Volumes mounted on a notebook.
  • Shared memory enabled:
  • Notebook creator: User who created the notebook.
  • Configurations: Configurations set when creating the notebook.
  • Type: Application used to run the notebook.
  • Minimum CPU: Minimum CPU value set at initial creation.
  • Maximum CPU: Maximum CPU value set at initial creation.
  • Minimum memory: Minimum memory value set at initial creation.
  • Maximum memory: Maximum memory value set at initial creation.
  • Image: Image selected when creating the notebook.
  • Environment: Runtime environment for the notebook.

Logs

You can view the recent logs for the selected notebook.

Events

You can view the history of recent events for the selected notebook.

  • Type: Type of the event that occurred.
  • Reason: Name of the event that occurred.
  • Created at: Date and time when the event occurred.
  • Message: Description of the event that occurred.

YAML

Shows the settings for the selected notebook in the YAML format.

Note

You can only view these details, but not edit them, from the console.