Hi @Phani1, Enabling customer-managed keys (CMKs) in Databricks on AWS involves securing and controlling access to encrypted data.
Here are the steps to set up CMKs:
Prerequisites:
- Ensure your Databricks workspace is on the E2 version of the Databricks Platform.
- You need to be on the Enterprise pricing tier to access these features.
Types of Data and Locations:
- Databricks has two customer-managed key features:
- Customer-managed keys for managed services: Used for various data types in the control plane.
- Customer-managed keys for workspace storage: Used for data stored in your workspace’s S3 bucket.
Data Types and Corresponding Key Features:
- Here’s a breakdown of which key feature to use for different types of data:
- Notebook source and metadata: Use customer-managed keys for managed services.
- Personal access tokens (PAT) or other credentials for Git integration: Use customer-managed keys for managed services.
- Secrets stored by the secret manager APIs: Use customer-managed keys for managed services.
- Databricks SQL queries and query history: Utilize customer-managed keys for managed services.
- Remote EBS volumes for Databricks Runtime cluster nodes: Apply customer-managed keys (applies only to classic compute plane resources).
- Workspace storage:
- Customer-accessible DBFS root data: Use customer-managed keys for workspace storage.
- Job results, Databricks SQL query results, MLflow Models, and Delta Live Table: All stored in your workspace’s S3 bucket, so use customer-managed keys for workspace storage.
- Interactive notebook results: Configure Databricks to store them in your AWS account using the appropriate key feature based on storage location.
Serverless Compute and Customer-Managed Keys:
- For Databricks SQL Serverless, you can use:
- Customer-managed keys for managed services for SQL queries and history.
- Customer-managed keys for your workspace’s S3 bucket (including root DBFS storage for SQL results).
- SQL warehouses do not use customer-managed keys for EBS storage encryption on compute nodes.
For a more detailed walkthrough, refer to the official Databricks documentation on customer-managed ....
🚀Our End-of-Year Community Survey is here! Please take a few moments to complete the survey. Your feedback matters!