Sneak: How to run a Direct Query

The Tips&Tricks Series is produced by the Customer Success Team. It highlights several easy to follow instructions and range from simple tasks to more advanced use cases.

How to run a direct query

Direct querying is powerful because it requires minimal setup and ensures that queries are built with the real-time data. As there is no data model output, there is no need to create any scheduling. Instead, whenever a user builds a query, Pyramid queries the actual data source, rather than data that was copied into Pyramid.
When setting up direct querying for a data source, additional nodes cannot be connected to a direct query data source node. Once you’ve selected this option you can continue directly to the Data Model part of the ETL, where you can construct the semantic data model.
Pyramid provides a list of built-in data sources and those that support direct query here.

1. Add the data source to be queried directly to the model data flow. Select the server and database then click the box “Direct Query Data Source”.
   Next select the tables that will be available in the direct query. Note that the “Add Tables” button is grayed out with the direct query option

   

2. Once the tables have been selected click on the “Data Model” tab and give the model a name. Then select the tables elements to confirm the joins between the tables, similarly the columns and Hierarchies elements can be adjusted as per the standard model functionality.

   

3. Process the model and the tables, and now the dimensions and measure included in the direct query are available in discover. The data is directly queried from the source database each time the discover is run.

   

If you have questions about these tips (or improvements) please let us know in a comment.

P.S. We will post previous Tip&Tricks as well as all new ones to keep this information active and available to all users.