Skip to main content

Local Agent Database Model

The local agent database model is a method of multidimensional data modeling using the database in the local agent as the data engine.

The database product used in the local agent is SQLite.

Creating a Local Agent Data Source​

If the user has a personal account, the system automatically creates a data source named Local - Agent DB of type LocalAgent on the data source page. Users can also create multiple data sources of type LocalAgent themselves.

The prerequisite for creating and using a data source of type LocalAgent is that the user has installed and connected to the Agent local agent service.

The data source creation interface is as follows:

  • Type: LocalAgent
  • Name: Unique technical name of the data source
  • Use Local Agent: Enabled, using the local agent database
  • Database: Filename of the SQLite database

Creating a Semantic Model​

In the semantic model management interface, create a new SQL model.

Select a data source of type LocalAgent

Choose the default default in the service directory, enter the unique technical name of the semantic model, and click apply to create.

Open the newly created semantic model, and click the upload button in the data table title bar to upload data.

The data file used in this article is AdventureWorks Sales.xlsx

In the data upload window, click or drag to upload data files in formats such as Excel, CSV, as shown below

Click next to preview the file data, change the name of each tab (i.e., table name), column names, and column types.

Click next to go to the upload page, click the delete button in front of each table row to remove tables you don't want to upload, and click upload to import data into the local agent database. Each table row will be imported one by one, and once the time in front of each row is completed, all will be successfully imported. The import time will be longer for large data volumes.

View table data and modeling, for detailed modeling functions, please refer to Semantic Modeling