Databricks

This guide is for Databricks. See also: Snowflake | BigQuery

Introduction

The ROX Databricks Integration enables customers to share live data directly from Databricks to ROX using Delta Sharing. Delta Sharing is an open protocol for secure data sharing that works across clouds and platforms—you retain full control while ROX gains real-time access to your datasets.

Key Benefits

ROX Account Information

Contact your ROX account team to obtain the ROX Metastore ID and Share name required for setting up Delta Sharing.

Your ROX account team will provide:

• Share name — The name to use when creating the share
• ROX Metastore ID — Required when creating the recipient

Setup Instructions

Step 1: Enable External Delta Sharing

1. In your Databricks workspace, navigate to Catalog → Delta Sharing → Share Data
2. If you see a warning that external sharing is not enabled, click Enable External Delta Sharing

1. Additionally, go to Account → Previews and enable Delta Sharing for Default Storage – Expanded Access

Step 2: Create the Share and Add ROX as Recipient

You can set up Delta Sharing via the UI or SQL.

Option 1: Using the UI

Create a Share:

1. Navigate to Catalog → Delta Sharing → Share Data
2. Click Create Share
3. Enter the share name provided by your ROX account team
4. Click Create

Add Data to the Share:

1. Click into your newly created share
2. Click Manage assets → Edit assets
3. Select the schemas and/or tables you want to share with ROX
4. Click Save

Add ROX as a Recipient:

1. Navigate to Catalog → Delta Sharing → Shared by me → Recipients
2. Click New recipient
3. Enter rox as the recipient name
4. Select Databricks-to-Databricks sharing
5. Enter the ROX Metastore ID provided by your ROX account team
6. Click Create

Grant Access to ROX:

1. Go back to your share (Catalog → Delta Sharing → Shared by me → Shares)
2. Click into your share
3. Click Manage recipients → Add recipients
4. Select the rox recipient
5. Click Add

Option 2: Using SQL

Open a SQL Editor in the Databricks workspace where your data lives and run the following commands:

-- 1. Create the share (share name will be provided by ROX)
CREATE SHARE IF NOT EXISTS <share_name_from_rox>;

-- 2. Add your schema(s) to the share
ALTER SHARE <share_name_from_rox> ADD SCHEMA <your_schema_name>;

-- 3. Create ROX as a recipient (Metastore ID will be provided by ROX)
CREATE RECIPIENT IF NOT EXISTS rox USING ID '<rox_metastore_id>';

-- 4. Grant ROX access to the share
GRANT SELECT ON SHARE <share_name_from_rox> TO RECIPIENT rox;

Tip: You can add multiple schemas by running additional ALTER SHARE ... ADD SCHEMA commands.

Complete SQL Example:

-- Replace placeholders with actual values from your ROX account team

-- 1. Create the share
CREATE SHARE IF NOT EXISTS ROX_CUSTOMER_SHARE;

-- 2. Add your schema(s)
ALTER SHARE ROX_CUSTOMER_SHARE ADD SCHEMA salesforce_data;
ALTER SHARE ROX_CUSTOMER_SHARE ADD SCHEMA crm_analytics;

-- 3. Create ROX as recipient
CREATE RECIPIENT IF NOT EXISTS rox USING ID 'abc123-your-rox-metastore-id';

-- 4. Grant access
GRANT SELECT ON SHARE ROX_CUSTOMER_SHARE TO RECIPIENT rox;

Step 3: Notify ROX

Once the share is configured, send the following to your ROX account team:

• Your Databricks workspace URL
• Share name
• Schema(s) and tables included in the share
• Brief description of shared data and its structure

Choosing What to Share

Best Practice: Share at the schema level when possible for simpler management. Use views to restrict access to sensitive columns or rows.

Information to Provide to ROX

After setting up the share, send the following to your ROX account team:

• Your Databricks workspace URL
• Share name
• Schema(s) and tables included in the share
• Brief description of shared data and its structure

Prerequisites

Technical Considerations

Security and Compliance

• All data sharing respects your existing Unity Catalog permissions
• ROX can only access schemas/tables explicitly added to the share
• Audit logs available in Unity Catalog for all access events

Performance

• Delta Sharing provides direct access to your Delta tables
• ROX queries run on ROX's compute resources, not yours
• No impact on your workspace's cluster performance

Data Freshness

• Real-time — ROX always queries your live Delta tables directly
• No replication delay or sync schedules

Troubleshooting