Separating operations and development when using user-managed notebooks: Use

This document describes how a data practitioner can use the notebooks manager and extended notebooks UIs for working with Vertex AI Workbench user-managed notebooks. It assumes that you are a data scientist, data analyst, or data engineer who is using a user-managed notebook that has been set up by an IT administrator as described in earlier documents in this series.

The document is part of a series that includes the following documents:

  • Overview, which describes the solution that you can use for deploying the notebooks manager and extended notebooks UIs.
  • Deploy, which guides IT administrators on how to deploy the notebooks manager and extended notebooks UIs.
  • Use (this document), which guides data practitioners on how to use the notebooks manager and extended notebooks UIs.
  • Troubleshooting, which describes potential issues and suggested resolutions.


Google Cloud customers can interact with Google products and services through clients and interfaces such as the console, REST API, client libraries, and the Google Cloud CLI. However, the console is the only visual client.

Use this tutorial if all of the following are true:

  • You do not have access to the console.
  • You use user-managed notebooks that are pre-created by IT administrators.
  • An IT administrator has given you a link to an alternative notebooks manager UI.

This document is intended for data practitioners and assumes the following:

  • You have access to user-managed notebooks environments and can start, stop, and list them.
  • Your identity has sufficient permissions to the relevant services that you can access them through the extended notebooks UI. BigQuery and Cloud Storage are some possible examples.
  • You understand how to use a Jupyter notebook and can run SQL queries.
  • You can write Python code.


  • Open the notebooks manager and extended notebooks UIs.
  • Start and stop user-managed notebooks instances.


This tutorial uses the following billable components of Google Cloud:

To generate a cost estimate based on your projected usage, use the pricing calculator. New Google Cloud users might be eligible for a free trial.

Before you begin

Make sure that you have a URL to access the notebooks manager page. You get this URL from your IT administrator. The URL is in the following form:

The URL includes a bucket name for BUCKET.

Access the notebooks manager

  1. From a web browser, access the notebooks manager using the URL that's provided by your IT administrator.

    If you're not already signed in, you must authenticate by completing the rest of this procedure. You need to sign in if you're using the web application for the first time, if you previously signed out of the UI, or if your OAuth 2.0 access is revoked.

  2. Click Sign In.

  3. Choose a Google account to authenticate with.

  4. Accept the different scopes that give the notebooks manager access to the APIs.

    For more information about scopes and APIs, see notebooks manager.

When you've signed in, you see the notebooks manager page.

Access the extended notebooks UI

You can now access the extended notebooks UI.

  1. In the notebooks manager page, verify that the user-managed notebooks instance is running, including the extended notebooks UI that you want to use.

    A green checkmark appears next to the instance name.

  2. Click Open JupyterLab.

    This assumes that the user-managed notebooks instance is running. If it isn't, the link is unavailable.

    Extended notebooks UI with a JupyterLab button.

    You can now interact with Google Cloud services through the extended notebooks UI.

    Interact with GCP services with the extended notebooks UI.

To bypass the notebooks manager, bookmark an extended notebooks UI URL.

Manage extended notebooks UI lifecycles

Actions on notebooks are limited by the signed-in user's IAM permissions. You can use the notebooks manager to manage the lifecycle of your extended notebooks UIs.

  1. To access an extended notebooks UI, verify that the underlying instance is running.

    If the instance is not running, start it by selecting the instance and clicking the Start button.

  2. To stop the instance, click Stop.

What's next