Trying to connect to DataBricks to pull data into Azure Synapse

Welcome to Microsoft Q&A

Hello @Steve McMasters I hope you are doing well,

What is actually happening is that Azure Synapse Analytics does not have a native Linked Service for the open Delta Sharing protocol. The standard Azure Databricks connector requires workspace and cluster compute permissions, which you do not have since you only possess an endpoint URL and a bearer token.

Check please the following things:

Use Synapse Spark: You must use a Serverless Apache Spark Pool in your Synapse workspace instead of a standard Copy Activity pipeline.

Install the Library: Add the open-source delta-sharing PyPI package to your Spark environment.

Run a Notebook: Create a config.share file with your URL and token, and use a PySpark notebook to read the data directly via delta_sharing.load_as_spark().

Follow this references from Microsoft Oficial Documentation this may help you to troubleshoot and resolve this issue.

References

Read data shared using Delta Sharing open sharing with bearer tokens

Copy data to and from Azure Databricks Delta Lake

If my answer helped you resolve your issue, please consider marking it as the correct answer. This helps others in the community find solutions more easily. Thanks

Hi ,

Thanks for reaching out to Microsoft Q&A.

you cannot directly use that Databricks Delta Sharing URL + bearer token as a native linked service in Azure Synapse today. Synapse does not support Delta Sharing as a first-class connector like PBI does.

What works instead is a workaround approach. The cleanest option is to use a compute layer that understands delta sharing (for example databricks, spark, or even a lightweight client) to pull the shared data and land it into your own storage such as ADLS Gen2 in Delta/Parquet, and then have Synapse read from there. If you want to stay closer to synapse, you can use Synapse Spark pools with the open-source delta sharing client (Python) to fetch data using the profile (URL + token) and write it into a lake database. Another option is orchestrating this via pipelines (ADF/synapse pipelines) calling a notebook or custom code activity.

PBI supports Delta Sharing natively, Synapse does not. You need an intermediate ingestion step to bridge that gap.

Please 'Upvote'(Thumbs-up) and 'Accept' as answer if the reply was helpful. This will be benefitting other community members who face the same issue.

Hi @Steve McMasters ,

Thank you for your question.

Based on your scenario, Azure Synapse does not currently support a native Linked Service to connect directly to a Delta Sharing server using only a URL and bearer token.

This is why your scenario works in Power BI — it includes a built-in Delta Sharing connector that supports authentication using URL and bearer token.

Recommended approach:

• Use a Synapse Spark (Serverless Spark pool) instead of Linked Service
• Use the Delta Sharing (open-source) library within Spark
• Access the data via a Spark notebook (PySpark)
• However, you can access the data using Synapse Spark with Delta Sharing as a supported workaround.

Microsoft Reference Links:

• Delta Sharing connector in Power BI (Microsoft Learn) [learn.microsoft.com]

Hope this helps. If you have any follow-up questions, please let me know. I would be happy to help.