Data sources supported in Azure Analysis Services

Data sources and connectors shown in Get Data or Import Wizard in Visual Studio are shown for both Azure Analysis Services and SQL Server Analysis Services. However, not all data sources and connectors shown are supported in Azure Analysis Services. The types of data sources you can connect to depend on many factors such as model compatibility level, available data connectors, authentication type, providers, and On-premises data gateway support.

Azure data sources

Datasource In-memory DirectQuery
Azure SQL Database2 Yes Yes
Azure SQL Data Warehouse Yes Yes
Azure Blob Storage1 Yes No
Azure Table Storage1 Yes No
Azure Cosmos DB1 Yes No
Azure Data Lake Store (Gen1)1, 4 Yes No
Azure HDInsight HDFS1 Yes No
Azure HDInsight Spark1, 3 Yes No

1 - Tabular 1400 and higher models only.
2 - Azure SQL Database Managed Instance is supported. Because a managed instance runs within Azure VNet with a private IP address, an On-premises Data Gateway is required. Azure SQL Database Managed Instance with a public endpoint is currently not supported.
3 - Azure Databricks using the Spark connector is currently not supported.
4 - ADLS Gen2 is currently not supported.

Provider
In-memory and DirectQuery models connecting to Azure data sources use .NET Framework Data Provider for SQL Server.

On-premises data sources

Connecting to on-premises data sources from and Azure AS server require an On-premises gateway. When using a gateway, 64-bit providers are required.

In-memory and DirectQuery

Datasource In-memory provider DirectQuery provider
SQL Server SQL Server Native Client 11.0, Microsoft OLE DB Provider for SQL Server, .NET Framework Data Provider for SQL Server .NET Framework Data Provider for SQL Server
SQL Server Data Warehouse SQL Server Native Client 11.0, Microsoft OLE DB Provider for SQL Server, .NET Framework Data Provider for SQL Server .NET Framework Data Provider for SQL Server
Oracle OLE DB Provider for Oracle, Oracle Data Provider for .NET Oracle Data Provider for .NET
Teradata OLE DB Provider for Teradata, Teradata Data Provider for .NET Teradata Data Provider for .NET

In-memory only

Datasource
Access Database
Active Directory1
Analysis Services
Analytics Platform System
CSV file
Dynamics CRM1
Excel workbook
Exchange1
Folder1
IBM Informix1 (Beta)
JSON document1
Lines from binary1
MySQL Database
OData Feed1
ODBC query
OLE DB
Postgre SQL Database1
Salesforce Objects1
Salesforce Reports1
SAP HANA1
SAP Business Warehouse1
SharePoint List1, 2
Sybase Database
TXT file
XML table1

1 - Tabular 1400 and higher models only.
2 - Files in on-premises SharePoint are not supported.

Specifying a different provider

Data models in Azure Analysis Services may require different data providers when connecting to certain data sources. In some cases, tabular models connecting to data sources using native providers such as SQL Server Native Client (SQLNCLI11) may return an error. If using native providers other than SQLOLEDB, you may see error message: The provider 'SQLNCLI11.1' is not registered. Or, if you have a DirectQuery model connecting to on-premises data sources and you use native providers, you may see error message: Error creating OLE DB row set. Incorrect syntax near 'LIMIT'.

When migrating an on-premises SQL Server Analysis Services tabular model to Azure Analysis Services, it may be necessary to change the provider.

To specify a provider

  1. In SSDT > Tabular Model Explorer > Data Sources, right-click a data source connection, and then click Edit Data Source.
  2. In Edit Connection, click Advanced to open the Advance properties window.
  3. In Set Advanced Properties > Providers, then select the appropriate provider.

Impersonation

In some cases, it may be necessary to specify a different impersonation account. Impersonation account can be specified in Visual Studio (SSDT) or SSMS.

For on-premises data sources:

  • If using SQL authentication, impersonation should be Service Account.
  • If using Windows authentication, set Windows user/password. For SQL Server, Windows authentication with a specific impersonation account is supported only for in-memory data models.

For cloud data sources:

  • If using SQL authentication, impersonation should be Service Account.

Next steps

On-premises gateway
Manage your server