A complete operational guide for managing Azure Databricks using the CLI covering authentication, workspaces, compute, jobs, Unity Catalog, secrets, DBFS, Git repos, and Infrastructure-as-Code via Asset Bundles.

Architectural Overview

The CLI communicates with Azure Databricks via REST APIs using OAuth-based authentication:

Local Machine (CLI)
        ↓
OAuth Authentication
        ↓
Azure Databricks Workspace
        ↓
Clusters / Jobs / Unity Catalog / Warehouses

1. Installation & Setup

Command	Description
winget --version	Verify Windows Package Manager is installed
winget install Databricks.DatabricksCLI	Install the Databricks CLI
databricks -v	Confirm installation and check version
where databricks	Show CLI binary path on disk

1.1 Check Winget

Confirm the Windows Package Manager is available before installing:

winget --version

1.2 Install Databricks CLI

winget install Databricks.DatabricksCLI

1.3 Verify Installation

databricks -v

Expected output: Databricks CLI v0.2xx.x

1.4 Check CLI Path

where databricks

2. Authentication (OAuth Based)

Command	Description
databricks auth login --host <url> --profile DEFAULT	Log in with OAuth (opens browser)
databricks auth describe --profile DEFAULT	Show active profile details
databricks auth profiles	List all configured profiles
Delete .databricks folder	Log out / remove cached tokens

2.1 Login

databricks auth login --host <workspace-url> --profile DEFAULT

Example:

databricks auth login --host https://adb-7405609716325528.8.azuredatabricks.net/

2.2 Check Profile

databricks auth describe --profile DEFAULT

2.3 List Profiles

databricks auth profiles

2.4 Logout

Delete the token cache folder to log out:

C:\Users\<username>\.databricks

3. Current User Information

databricks current-user me

Returns: username, email address, and user ID.

4. Workspace Commands

Command	Description
databricks workspace list /	List root workspace contents
databricks workspace mkdirs /path	Create a folder
databricks workspace delete /path	Delete a file
databricks workspace delete /path --recursive	Recursively delete a folder
databricks workspace import /dest --file local.py --language PYTHON --format SOURCE	Upload a file
databricks workspace export /src --file out.py --format SOURCE	Download a file

4.1 List Workspace Root

databricks workspace list /

4.2 Create Folder

databricks workspace mkdirs /Shared/Yash

4.3 Delete File or Folder

Delete a single file:

databricks workspace delete /Shared/Yash/file.py

Recursive delete:

databricks workspace delete /Shared/Yash --recursive

4.4 Import File

databricks workspace import /Shared/Yash/local.py \
--file local.py --language PYTHON --format SOURCE --overwrite

Verify upload:

databricks workspace list /Shared/Yash

4.5 Export File

databricks workspace export /Shared/Yash/local.py \
--file exported.py --format SOURCE

Verify locally:

dir exported.py

4.6 List Files

databricks workspace list /Shared/Yash

5. Cluster Management

Command	Description
databricks clusters list	List all clusters
databricks clusters get <cluster-id>	Get full cluster details/config
databricks clusters start <cluster-id>	Start a terminated cluster
databricks clusters delete <cluster-id>	Terminate (stop) a running cluster
databricks clusters permanent-delete <cluster-id>	Permanently remove a cluster

5.1 List Clusters

databricks clusters list

5.2 Get Cluster Details

databricks clusters get 0106-075344-c5jhxigd

5.3 Start Cluster

databricks clusters start 0106-075344-c5jhxigd

5.4 Terminate Cluster

databricks clusters delete 0106-075344-c5jhxigd

5.5 Permanent Delete

databricks clusters permanent-delete <cluster-id>

6. Jobs Management

Command	Description
databricks jobs list	List all jobs
databricks jobs get <job-id>	Get job configuration and details
databricks jobs run-now <job-id>	Trigger an immediate run
databricks jobs list-runs	List recent job runs and statuses
databricks jobs cancel-run <run-id>	Cancel an active run
databricks jobs delete <job-id>	Delete a job definition

6.1 List Jobs

databricks jobs list

6.2 Get Job Details

databricks jobs get 909703333311409

6.3 Run Job Now

databricks jobs run-now 909703333311409

6.4 List Runs

databricks jobs list-runs

6.5 Cancel Run

databricks jobs cancel-run 909703333311409

6.6 Delete Job

databricks jobs delete 909703333311409

7. SQL Warehouses

Command	Description
databricks warehouses list	List all SQL warehouses
databricks warehouses start <id>	Start a stopped warehouse
databricks warehouses stop <id>	Stop a running warehouse

7.1 List Warehouses

databricks warehouses list

7.2 Start Warehouse

databricks warehouses start dd2325bf3cbab4f6

7.3 Stop Warehouse

databricks warehouses stop dd2325bf3cbab4f6

8. Unity Catalog Commands

Command	Description
databricks catalogs list	List all catalogs
databricks catalogs get <catalog>	Get catalog metadata
databricks schemas list <catalog>	List schemas in a catalog
databricks tables list <catalog> <schema>	List tables in a schema
databricks tables get <catalog>.<schema>.<table>	Get full table details

8.1 List Catalogs

databricks catalogs list

8.2 Get Catalog

databricks catalogs get fq_dev_catalog

8.3 List Schemas

databricks schemas list fq_dev_catalog

8.4 List Tables

databricks tables list <CATALOG_NAME> <SCHEMA_NAME>

Example:

databricks tables list fq_dev_catalog gold

8.5 Get Table Details

databricks tables get <CATALOG>.<SCHEMA>.<TABLE>

Example:

databricks tables get fq_dev_catalog.gold.sales_sample

9. Secrets Management

Command	Description
databricks secrets create-scope <scope>	Create a new secret scope
databricks secrets list-scopes	List all secret scopes
databricks secrets put-secret <scope> <key>	Add or update a secret (prompted input)
databricks secrets list-secrets <scope>	List key names in a scope
databricks secrets delete-secret <scope> <key>	Delete a specific secret key
databricks secrets delete-scope <scope>	Delete entire scope + all secrets

9.1 Create Secret Scope

databricks secrets create-scope my-scope

9.2 List All Secret Scopes

databricks secrets list-scopes

9.3 Add Secret

databricks secrets put-secret my-scope my-key

You will be prompted (input is hidden):

Enter secret value: ****

9.4 List Secrets in Scope

databricks secrets list-secrets my-scope

Output:

Key
my-key

9.5 Delete a Secret

databricks secrets delete-secret my-scope my-key

9.6 Delete Entire Scope

databricks secrets delete-scope my-scope

10. Repos (Git Integration)

Command	Description
databricks repos list	List connected Git repos
databricks repos create --url <git-url> --provider github	Connect a new repo
databricks repos delete <repo-id>	Disconnect a repo

10.1 List Repos

databricks repos list

10.2 Create Repo

databricks repos create --url https://github.com/repo.git --provider github

10.3 Delete Repo

databricks repos delete <repo-id>

11. Files (DBFS)

Command	Description
databricks fs ls dbfs:/	List root DBFS contents
databricks fs cp local.csv dbfs:/tmp/	Copy a file to DBFS
databricks fs ls dbfs:/tmp/	Verify uploaded file
databricks fs rm dbfs:/tmp/local.csv	Remove a file from DBFS

11.1 List DBFS

databricks fs ls dbfs:/

11.2 Copy File to DBFS

databricks fs cp local.csv dbfs:/tmp/local.csv

11.3 Verify Upload

databricks fs ls dbfs:/tmp/

11.4 Remove File

databricks fs rm dbfs:/tmp/local.csv

11.5 Verify Deletion

databricks fs ls dbfs:/tmp/

12. Databricks Asset Bundles (DAB)

Asset Bundles enable Infrastructure-as-Code (IaC) — deploy jobs, pipelines, and resources in a repeatable, version-controlled way.

Command	Description
databricks bundle init	Scaffold a new bundle project
databricks bundle validate	Validate bundle configuration
databricks bundle deploy -t dev	Deploy bundle to the dev target
databricks bundle destroy -t dev	Tear down a deployed bundle

12.1 Initialize Project

mkdir my_dab_project
cd my_dab_project
databricks bundle init

12.2 Validate Bundle

databricks bundle validate

12.3 Deploy Bundle

databricks bundle deploy -t dev

12.4 Destroy Deployment

databricks bundle destroy -t dev

13. Debug & Advanced

Enable Debug Logs

databricks --debug clusters list

Check CLI Help

databricks --help

Command-specific:

databricks clusters --help

14. Profile Management (Multiple Environments)

Use named profiles to switch between Dev, QA, and Prod workspaces without re-authenticating each time.

Command	Description
databricks auth login --host <url> --profile PROD	Add a named profile
databricks clusters list --profile PROD	Run any command against a profile
databricks auth profiles	List all saved profiles

Login with New Profile

databricks auth login --host <url> --profile PROD

Use a Specific Profile

databricks clusters list --profile PROD

Real Use Cases

• Automating cluster lifecycle — start/stop/terminate on a schedule via cron or CI/CD
• Deploying jobs via CI/CD pipelines (GitHub Actions, Azure DevOps, Jenkins)
• Managing Unity Catalog — create schemas, manage table permissions at scale
• Bulk workspace operations — batch import/export of notebooks across projects
• Infrastructure as Code — version-controlled deployments with Asset Bundles

Security & Best Practices

• Prefer OAuth authentication it supports automatic token refresh
• Use profiles to isolate Dev / QA / Prod environments
• Never hardcode cluster IDs, workspace URLs, or secrets in scripts
• Use Asset Bundles for repeatable, peer-reviewed deployments
• Store credentials in Databricks Secrets never in plain text or source repos
• Delete the .databricks cache folder when decommissioning a machine