Introduction to managed notebooks

Vertex AI Workbench managed notebooks isdeprecated. On April 14, 2025, support for managed notebooks ended and the ability to create managed notebooks instances was removed. Existing instances will continue to function until March 30, 2026, but patches, updates, and upgrades won't be available. To continue using Vertex AI Workbench, we recommend that youmigrate your managed notebooks instances to Vertex AI Workbench instances.

Vertex AI Workbench managed notebooks instancesare Google-managed environmentswith integrations and capabilities that help you set up and work inan end-to-end Jupyter notebook-based production environment.

Managed notebooks instances are prepackaged withJupyterLaband have a preinstalled suite of deep learning packages,including support for the TensorFlow and PyTorchframeworks. Managed notebooks instances support GPU accelerators andthe ability to sync with aGitHub repository.Your managed notebooks instances are protectedby Google Cloud authentication and authorization.

Google-managed compute infrastructure

A Vertex AI Workbench managed notebooks instanceis a Google-managed, Jupyter notebook-based, compute infrastructure.

When you create a managed notebooks instance,it is deployed as a Google-managed virtual machine (VM) instance in atenant project.

Your managed notebooks instance includes many commondata science framework environments, such as TensorFlowand PyTorch. You can also add your own custom container images toyour managed notebooks instance. These environmentsare available as kernels that you can run yournotebook file in.

When you run a notebook in one of the kernels, Vertex AI Workbenchstarts the corresponding container, creates a Jupyter session on it, anduses that Jupyter session to run your notebook on the container.

This Google-managed compute infrastructure includes integrationsand capabilities that help you implement data science and machine learningworkflows from start to finish. See the following sections for details.

Use custom containers

You can add custom Docker container images toyour managed notebooks instanceto run your notebook code in an environment customized for your needs.

These custom containers are available to use directly from theJupyterLab user interface, alongside the preinstalled frameworks.For more information, seeAdd a custom container toa managed notebooks instance.

Notebook-based workflow

Managed notebooks instances let youperform workflow-oriented tasks without leaving the JupyterLab user interface.

Control your hardware and framework from JupyterLab

In a managed notebooks instance, your JupyterLab user interfaceis where you specify what compute resources your code will run on. For example,you can configure how many vCPUs or GPUs you want, how much RAM you want, andwhat framework you want to run the code in. You can write your code first, andthen choose how to run it without leaving JupyterLabor restarting your instance.For quick tests of your code, you can scale your hardware down and then scale itback up to run your code against more data.

Access to data

You can access your data without leaving the JupyterLab user interface.

In JupyterLab's navigation menu ona managed notebooks instance, you can use theCloud Storage integrationto browse data and other files that you have access to.SeeAccess Cloud Storage buckets and filesfrom within JupyterLab.

You can also use theBigQuery integrationto browse tables that you have access to,write queries, preview results, and load data into your notebook.SeeQuery data in BigQuery tablesfrom within JupyterLab.

Execute notebook runs

Use the executor to run a notebook fileas a one-time execution or on a schedule.Choose the specific environment and hardware that you wantyour execution to run on. Your notebook's code will run onVertex AI custom training, which can make it easierto do distributed training, optimize hyperparameters, orschedule continuous training jobs. SeeRun notebook fileswith the executor.

You canuse parameters inyour executionto make specific changes to each run.For example, you might specify a different dataset to use,change the learning rate on your model, or change the versionof the model.

You can alsoset a notebook to run on a recurringschedule.Even while your instance is shut down, Vertex AI Workbench willrun your notebook file and save the resultsfor you to look at and share with others.

Share insights

Executed notebook runs are stored in a Cloud Storage bucket,so you can share your insights with others by granting accessto the results. See theprevious section on executingnotebook runs.

Secure your instance

You can deploy your managed notebooks instancewith the default Google-managed network,which uses a default VPC network and subnet.Instead of the default network, you can specify aVPC network to use with your instance.For more information, seeSet up a network. You can useVPC Service Controlsto provide additional security for yourmanaged notebooks instances.

To use managed notebooks within a service perimeter, seeUsea managed notebooks instance within a serviceperimeter.

By default, Google Cloud automaticallyencrypts data when it is atrestusing encryption keys managed by Google. If you have specific compliance orregulatory requirements related to the keys that protect your data, you canuse customer-managed encryption keys (CMEK) withyour managed notebooks instances. For more information,seeUse customer-managed encryption keys.

Automated shutdown for idle instances

To help manage costs,managed notebooks instancesshut down after being idle for a specific time period by default.You can change the amount of time or turn this feature off.For more information,seeIdle shutdown.

Dataproc integration

You can process data quickly by running a notebookon a Dataproc cluster.After your cluster is set up, you can runa notebook file on it without leaving the JupyterLab user interface.For more information, seeRun a managed notebooks instanceon a Dataproc cluster.

Limitations

Consider the following limitations ofmanaged notebooks when planning your project:

• Managed notebooks instances are Google-managedand therefore less customizable than Vertex AI Workbenchuser-managed notebooks instances.User-managed notebooks instances can bemore ideal for users who need a lot of control over their environment.For more information, seeIntroduction touser-managed notebooks.

• Third party JupyterLab extensions are not supported.

• The Dataproc JupyterLab plugin isn't supported formanaged notebooks, but you can use the plugin inVertex AI Workbench instances. SeeCreate aDataproc-enabledinstance.

• Managed notebooks instances do not allow users tohavesudo access.

• When you useAccess Context ManagerandChrome Enterprise Premiumto protect managed notebooks instances withcontext-aware access controls, access is evaluated each timethe user authenticates to the instance. For example, accessis evaluated the first time the user accesses JupyterLab andwhenever they access it thereafter if their web browser'scookie has expired.

• To use accelerators with managed notebooks instances,the accelerator type that you want must be available in your instance'szone. To learn about accelerator availability by zone, seeGPU regions and zones availability.

What's next

• Learn more about thenetworking options available for yourmanaged notebooks instance.