Skip to main content
Version: 9.3

Synapse Configuration

To ensure proper communication between Qrvey and Azure Synapse, complete the following configuration tasks after the initial Qrvey deployment. By default, a new deployment creates the Synapse workspace resource.

Before You Begin​

  1. From the Azure CLI, enable the Synapse service:

    az provider register --namespace Microsoft.Synapse

    From the Azure web console, you can also enable (register) the provider in your subscription settings.

  2. Identify the managed identity.

    Navigate to the resource group where Qrvey was deployed. Locate the Managed Identity resource (for example, lyocmmaz-synapse-wi) used to run SQL scripts.

  3. Acquire administrator permissions:

    • Contributor or Owner permissions on the Synapse workspace resource to assign RBAC roles.
    • Synapse Administrator permissions to execute scripts against the SQL pools.

Step 1: Assign Synapse RBAC Roles​

Before running the SQL scripts, the administrative user must ensure he or she has the necessary access within Synapse Studio:

  1. In the Azure Portal, open your Synapse Workspace.
  2. Navigate to Manage > Access Control and verify that the user executing the scripts is assigned the Synapse Administrator role.

Step 2: Execute SQL Configuration Scripts​

  1. Open Synapse Studio and navigate to the Develop tab.

  2. Create a new SQL script. Ensure you are connected to the Built-in pool.

    a. Set up the master database. Select the master database from the Use database dropdown and run the following SQL commands:

    CREATE LOGIN [<managed-identity-name>] FROM EXTERNAL PROVIDER;
    ALTER SERVER ROLE sysadmin ADD MEMBER [<managed-identity-name>];

    b. Set up the application database. Select your specific application database (for example, qrveydb_) from the dropdown and run the following SQL commands:

    CREATE USER <managed-identity-name> FROM EXTERNAL PROVIDER;
    ALTER ROLE db_owner ADD MEMBER [<managed-identity-name>];

Joined Data​

When multiple data sources are joined, text columns are truncated to 8000 characters. This restriction does not apply to datasets that use a single data source.