Importing InterSystems IRIS Business Intelligence Models

AtScale supports conversion of InterSystems IRIS Business Intelligence models to AtScale models. The importing process applies best practices for converting InterSystems IRIS Business Intelligence semantics to AtScale semantics. However, the imported model must be reviewed and tested to ensure desired operation.

Although the semantics of the two model types are similar, they do not match exactly. AtScale will not convert InterSystems IRIS model components that are not supported in AtScale. Additionally, some InterSystems IRIS modeling concepts are supported in a different way in AtScale. In these cases, the importer will create the equivalent AtScale model structure. The import report contains a comprehensive record of the conversion process, including items omitted because they are unsupported. For details, see the "Limitations of the AtScale InterSystems IRIS Importer" section below.

Before you begin

  • Ensure that your AtScale instance has a license key that will enable InterSystems IRIS support.

  • Ensure that your user ID is assigned to a role that has the Create Projects permission for the current organization. By default, the Organization Admin role has this permission.

  • Configure an AtScale Data Warehouse Connection to an IRIS data warehouse that stores the tables you wish to access from AtScale. For details, see Adding InterSystems IRIS Data Warehouses.

  • Export the InterSystems IRIS database definition in XML format.

    Consider that the InterSystems IRIS model exporter functionality is available in IRIS 2021.2 and later versions.

Procedure

  1. On the AtScale home page's Quick Start section, locate the Import from InterSystems IRIS option and choose it.

  2. On the Import InterSystems IRIS file step in the Project Wizard, choose the Browse button, select the corresponding IRIS export .json file, and choose Next to continue.

    In case AtScale detects errors in the file, you will see an error message. If this happens, you have to correct any errors in the file and re-upload it before proceeding.

  3. On the Select Data Warehouse and Schema step:

    1. Enter the project name. The default value is the project name from the imported file.

    2. Select data warehouse.

    3. For each dataset, select a schema (schemas specified in the project file are automatically selected).

    4. Choose Next.

      You may have to wait while the system processes the project file. The wait time depends on the size of the model and network bandwidth. Do not click any buttons while waiting.

  4. On the Review Import step:

    1. Check all messages in the import summary.

    2. Download and review the conversion report. It contains more details about the conversion process in the form of info, warning, and error messages.

    3. Do one of the following

      • Choose the Back button to change the project, data warehouse, or schema.
      • Choose the Next button to continue.
  5. On the Set Permissions step, set the initial permissions for your project, and choose Next to continue.

    Projects and cubes are created with the system-configured default design-time and run-time security values. The default system values for design-time and run-time access are open to all users. A Super User Administrator can change this behavior to Exclusive Access by changing the Default Project/Cube Security from the Settings > Organization Settings: Options page.

    Note: A user must have the Administer Organization permission, or be the installation Administrator to change run-time and design-time security settings. When creating a project with a Wizard, a non-administrator can see the security settings of the project, but cannot change them.

    Choose one of the following:

    • Yes, allow all: Allow all users in the organization to read, update, delete, and publish the project. The initial cube in the project is created and all users in the organization have read, update, and delete permissions on it.
    • No, just me: Allow only yourself, organization admins for the current organization, and super users to access the project. The initial cube in the project is created with the same restriction.
  6. Click Finish to exit the wizard and load the cube.

Post-Procedure

  1. If the data warehouse does not contain the same schemas and tables as the InterSystems IRIS database you must do one of the following:

    • Migrate the InterSystems IRIS schema and tables to the data warehouse.
    • Update each dataset in the AtScale model with the desired schema and table names.
  2. Review the import report and make any additional edits to the model to finish your InterSystems IRIS migration. Common activities include:

    • Review the Suggestions and Warnings that are created during the import process. This information may be saved as text file from the Import Wizard.
    • Testing the automatically created Degenerate Dimensions with your BI tool.
    • Features that are not supported by the importer should be added manually to the model, as described in 'Limitations of the AtScale InterSystems IRIS Importer' below.
    • Migrate unsupported MDX calculations to either the AtScale cube as rephrased calculated measures, calculated columns, query datasets, or to workbook calculations.
    • Make non-unique captions unique. InterSystems IRIS allows for duplicate attribute caption names, however AtScale discourages this practice. AtScale will display the following warning message on attribute dialogs if duplicate attribute names are detected, "This display name is already in use". To avoid accidentally using the wrong attribute in reports and calculations you should only use globally unique attribute captions.
  3. Test the model with Cube Data Preview.

  4. Publish the model, and query it with a supported Business Intelligence Tool.

Limitations of the AtScale InterSystems IRIS Importer

  • InterSystems IRIS Features that Require Manual Creation in AtScale

    Create the following features in AtScale with Design Center:

  • Differently Supported

    The following InterSystems IRIS semantics are converted to AtScale semantics by the importer.

    • Time Dimensions that are defined from the Fact Dataset must have at least one record for each Time Member. Alternatively, if the Fact Dataset cannot satisfy this requirement, then the Time Dimension can be manually modeled to reference a normalized Time table that you create in your IRIS data warehouse. The normalized Time table must have one record for each time member in your date dimension. The Fact Dataset should have a foreign-key relationship with the Time table.