Checklist For Creating a New Analysis Services Project

March 30, 2019

Setting up a new SSAS Multidimensional project in SSDT is pretty straight forward. There’s a few things to do before creating dimensions and measure groups. Here’s my checklist:

  • Set Project Properties – Right click the project name and click properties
    • Deployment Server Edition – This property is on the build tab of the property page. By default this is set to Developer. Developer enables every feature available in SSAS Enterprise edition. It might be a good idea to match the version that you’re licensed for. This way, you don’t use a feature in development that’s not available in your production environment.
    • Server – On the deployment tab of the properties page. This is the name of the SSAS server you will be deploying to.
    • Database – On the deployment tab of the properties page. This is the name of the Database you’re deploying to.
    • Processing Option – On the deployment tab of the properties page. Analysis Services needs to process any object that’s changed. This processing can occur as part of the deployment process. In fact, this is the default setting for Processing Option. It’s also possible to deploy without processing affected objects. This is probably the better option. For example, objects with a lot of data might take a while to process. To avoid processing during deployment change this property value to Do Not Process.
  • Create Data Source
    • Configure the Data Source – If the data source is SQL Server use the native client over the .net providers.
    • Set the correct Impersonation Information – This controls how SSAS connects back to the Data Source. The options are:
      • Windows user name and password – Provide a windows user name and password.
      • Service account – This will attemt to use the account the SSAS Service is running on. If SSAS is using a built in account or NetworkService, a database login must exist and have read permissions.
      • credentials of current user – This only works for local cubes. Probably not a good option.
      • Inherit – This option uses the
        Data Source Impersonation Info option set at the database level.
  • Create Data Source View
    • Create Diagrams using the Diagram Organizer – This is a neat way to organize measure groups. The Diagram Organizer is in the top left corner of the Data Source View designer.
    • Ensure the relationships are correct – The more work that’s done here the easier it is to develop dimensions and measure groups later.