Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Important
Databricks Asset Bundles in the workspace is in Public Preview.
Databricks Asset Bundles are created, deployed to a workspace, then managed programmatically from your local development environment or directly in the workspace UI. Collaborating on bundles directly in the workspace allows for more rapid iteration and testing before moving to production.
For more information about bundles, see What are Databricks Asset Bundles?.
Tip
If you don't want to use the workspace UI to create and manage your bundles, but would still like to develop bundles in the workspace, you can use the Databricks CLI from the workspace web terminal, just as you would from your local terminal. See Run Databricks CLI commands.
What are the installation requirements of bundles in the workspace?
You do not have to install anything locally to use bundles in the workspace, but there are Databricks workspace requirements:
- Workspace files must be enabled. See What are workspace files?.
- You must have a Git folder in which to create the bundle. To create a Git folder, see Clone a repo connected to a remote Git repository.
- Serverless compute must be enabled. See Enable serverless compute.
The version of the Databricks CLI used to deploy a bundle and run resources in the workspace is provided in the Deploy dialog.
Do I have to understand how to author YAML to use bundles in the workspace?
No. Databricks Asset Bundles in the workspace provides a UI for managing bundles so that you can collaborate on bundles that were developed by other users in their local development environments. You do not need to learn YAML or know how to use the Databricks CLI to work with bundles in the workspace.
I already manage my workflows in the Databricks UI. Why should I use bundles?
Databricks Asset Bundles allows you to source control all of the files needed for your workflows. Bundles in the workspace provides a mechanism for workspace users to edit, commit, test, and deploy updates through the UI. Users launch a bundle project from within a Git folder.
I have a bundle in a GitHub repository. How can I edit it in the Databricks workspace?
A folder is identified as a bundle by Databricks if a databricks.yml file exists at the root of the folder. Create a Git folder in Databricks for your GitHub repository, and Databricks will recognize it as a bundle.
How do I share a bundle for collaboration?
Bundles inherit the permissions of their parent Git folder, so to share a bundle, share the Git folder that contains your bundle project with your collaborators. See Collaborate, review, and deploy to production.
How can I move a bundle to production in the workspace?
Bundles define a complete project to deploy, the infrastructure necessary to run the project, and the workspaces to target for deployments. These target workspaces are typically your development, staging and production environments. Databricks Asset Bundles in the workspace makes it easy to switch targets, deploy, and run workflows from the UI. See Deploy bundles and run workflows from the workspace.