One option is using Github for version control. You can link a notebook and Databricks syncs your history with Git- which gives you the option to restore an older version if needed. Here are the docs for more details: https://docs.databricks.com/notebooks/github-version-control.html#github-enterpriseebooks/github-version-control.html#github-enterprise
You can use version controlled source code for you databricks job and each time you need to rollback to older version of your job you need just to move to older version code. For version controlled source code you have multiple choises: - Use a notebook from a remote Git repository - Use Python code from a remote Git repository - Use SQL queries from a remote Git repository https://docs.databricks.com/en/workflows/jobs/how-to/use-repos.html