Posted by Guest
Aug 29, 2016

Guest post by Senior Program Manager for Power BI at Microsoft. This post can also be found on the Power BI Blog.

As part of the Power BI Desktop August Update we are very excited to announce a preview of a new data connector for Snowflake.

Snowflake is a cloud-native elastic data warehouse service that makes it easy to bring together data from disparate sources and make it available to all users and systems that need to analyze it. With the new Power BI Desktop connector, users can import data from Snowflake, combine it with data from other sources, and build rich data visuals as part of their Power BI reports.

In order to use this new connector, users need to first enable the Preview feature option within the Options dialog (File -> Options and Settings -> Options).

clip_image001

As a pre-requirement, the Snowflake connector requires users to install the Snowflake ODBC driver in their machines, matching the architecture of the Power BI Desktop installation (i.e. 32-bit vs. 64-bit). The Snowflake ODBC driver can be downloaded from this location: http://go.microsoft.com/fwlink/?LinkID=823762

After the feature has been enabled, users can find the Snowflake connector within the “Get Data” dialog, under the Database category.

clip_image002

In the connector dialog, users can specify the Snowflake Server and Warehouse to connect to.

clip_image004

Once the connection details have been specified, Power BI Desktop will attempt to connect to Snowflake, which will require users to specify credentials. In Power BI Desktop, credentials are only asked from the user on first-time connection, after which they are stored as part of the user profile and can be managed from the “Data Source Settings” dialog (under File -> Options and Settings). Subsequent connections to Snowflake will reuse the stored credentials for the current user.

clip_image006

After the connection has been successfully established and the user has been authenticated, the Navigator dialog will allow users to browse the list of databases available within the specified warehouse. In this dialog, users can preview and select one or multiple tables to use in their Power BI report.

clip_image008
Once one or more tables have been selected in the Navigator dialog, users can either:

  • Load: This will load all data rows from each of the tables selected into the local Power BI Desktop data model, from where users can build their reports.
  • Edit: This option will bring users into the Query Editor dialog, from where they can perform data transformations and filters on a sample subset of the data, without the need to download the entire data set. Operations applied within the Query Editor dialog will run on the Snowflake backend, as long as the operations are supported in the source (i.e. filters, etc.)

We’re very glad to enable connectivity to Snowflake from Power BI and are looking forward to customers’ feedback. We encourage you to give it a try and share with us any feedback or issues that you encounter via the “Send a Frown” feature in Power BI Desktop. In subsequent updates, we will also enable DirectQuery capabilities for this connector, as well as support for refreshing datasets based on Snowflake data in the Power BI service via the Power BI Data Gateway…

Stay tuned for improvements to the Snowflake connector and for many other connectors in upcoming releases. If yours isn’t supported yet, just let us know.

Miguel Llopis