Model Data in Snowflake Data Warehouse

Configure your on-premises environment so that you can model data in a Snowflake database.

You can model Snowflake data in Semantic Modeler or Model Administration Tool. This task describes the process using Model Administration Tool.
Both local and remote connections to Snowflake from a semantic model require an installation of Data Gateway in order for the Model Administration Tool to import and model tables from a Snowflake data source. Configure the Data Gateway agent with a suitable driver to connect to Snowflake. Make sure that remote Data Gateway connections are available when queries are run. However, once the Snowflake tables are modeled and the semantic model is published to Oracle Analytics, for local connections, Data Gateway can be disabled or removed as it isn't used when queries are run from Oracle Analytics.

You can create either a local or a remote connection to Snowflake from the semantic model. When creating a local (not remote) connection, the connection pool in the semantic model uses a JDBC connection.

Before you start, install Data Gateway and Model Administration Tool on the same Windows computer in your on-premises environment.
  1. Configure a local Data Gateway agent to facilitate a connection from the Developer Client Tool to Snowflake.
    1. Download the latest Snowflake JDBC driver (for example, in file snowflake-jdbc-3.9.0.jar).
    2. Copy the downloaded JAR file to the Data Gateway installation folder.
      In a server deployment, copy the JAR file into:
      <Data Gateway folder>/domain/jettybase/lib/ext
      In a personal deployment, copy the JAR file into:
      <install directory>\war\datagateway\WEB-INF\lib
    3. Re-start Data Gateway.
  2. Configure a connection to your Snowflake database.