External Data Sharing With Microsoft Fabric

Organizations leveraging Microsoft Fabric can now use the external data sharing feature to allow users from another Microsoft Fabric tenant access their data. The data stays in the provider's OneLake storage location and no data is moved. Instead, this cross-tenant sharing creates a link, called a OneLake shortcut, and the recipient of the share tenant creates a pointer back to the provider tenant. Users in the recipient tenant are provided a read only access to the data. The data is shared across tenant boundaries, and the recipient can use it with any OneLake compatible Fabric workload in their own tenant. What makes external sharing solution so seamless and frictionless is the fact that the adoption of PowerBI already has massive adoption in terms of the number of organizations and the millions of users who leverage it daily. The key becomes, organizations don't need to do anything different nor stand up any new products and solutions; organizations and users can simply continue to use what they have been using and take advantage of this new feature.

External data sharing is currently supported for data residing in tables or files within:

• Lakehouses
• KQL databases

We wanted to provide some high level step by step guidance to demonstrate how easy it was to setup the external sharing:

Step 1 Enable External Tenant Sharing:

1. Enable external data sharing in the provider tenant

1. Go to the tenant settings in the providing tenant.
2. Find the External data sharing (preview) tenant setting (under export and sharing settings), enable the toggle, and specify who in the tenant can create external data shares.
3. Decide if you want to enable this capability for everyone or limit to certain security groups.

Step 2 Create an external data share:

1. In a workspace or in the OneLake data hub, find the Fabric item that contains the data you want to share.
2. Open the context menu of the selected item and choose External data share (preview).
3. In the New external data share dialog that opens, choose the folder or table to be shared and select Save and continue. 

Enter the email address of the user you want to share the data with and select Send. An email will be sent to the user inviting them to accept the invitation to share. Alternatively, you can select Copy link and then paste it into an email that you yourself compose. 

Note – it is just as easy to revoke the share – under manage permissions:

Step 3 Allow/accept external data share in recipient tenant:

1. Go to the tenant settings in the recipient tenant.
2. Find the Users can accept external data shares (preview) tenant setting (under export and sharing settings), enable the toggle, and specify who in the tenant can create external data shares.
3. Decide if you want to enable this capability for everyone or limit to certain security groups 

Accept an external data share:

1. Go to the link received via email and click on Get Started. 

2. A dialog similar to one below appears that displays the name of the share and the details of the data provider's tenant. 

3. Select Accept and select a location. This opens the OneLake data hub. Select a lakehouse. 

4. Select Next, select the table or folder location in which to create the incoming share shortcut, and then select Apply. 

Observations:

1. The external data share shows up as a shortcut with the name starting with a “External_”. 
2. The external data share gives a read-only version of data that can be accessed through all the Fabric experiences.  For the spark notebook experience, users can choose from multiple languages or the low code/no code data wrangler experience to transform and prepare the data. In the example below, the table “store” is combined with the external shortcut “EXTERNAL_storeprofile”, which provides the store profile data, to get a complete view of the store features. The result is then saved to a new table, “storewprofile”. This new table can be accessed through the SQL Analytics Endpoint, Fabric Data Factory, or PBI experiences. 

Here is an example of the SQL Analytics endpoint on the "External_" share:

3. The external share can be viewed through the lineage view of the workspace. 

Key Takeaways 

The external data share is a very powerful feature and so we hope you take away the following from the content we have shared: 

1. Ease of creating and revoking a share. 
2. Simplicity of accepting a share as the recipient. 
3. Fabric Ecosystem & Non-Fabric Use Cases – although we have demonstrated a very Fabric centric article, non-fabric users can still benefit from this solution. For instance, if you are a Tableau user, you can connect to the XMLA endpoint of a semantic model or the SQL endpoint of Fabric via the necessary drivers.  Given this, the solution only becomes even more compelling as it can handle both Fabric centric and non Fabric centric scenarios.