In a semantic model deployed to Development, Test, and Production, to reduce the size of query requests during full semantic model refreshes in Development or Test, what should you do in the deployment pipeline?

• A. Add a deployment parameter rule to filter the data.
• B. Configure row-level security (RLS).
• C. Connect either workspace to an Azure Data Lake Storage Gen2 account.
• D. Enable an incremental refresh policy.

Answer: (A)

Using deployment parameter rules to filter data lets you tailor what gets loaded in each environment. By applying a deployment parameter rule in Development and Test, you limit the data included in the semantic model during a refresh, so the model is smaller and the refresh requests are lighter. This directly reduces the amount of data processed and transferred, speeding up refreshes and saving resources in non-production environments, while Production can be left with the full dataset.

Row-level security restricts data after loading for individual users, so it doesn’t shrink the data loaded during a full refresh. Connecting the workspace to Azure Data Lake Storage Gen2 changes where data is sourced, not how much data is refreshed. An incremental refresh policy helps only if you’re performing incremental loads; during a full refresh, it wouldn’t reduce the size of the refresh payload.