Knowledge Check

1.

A data engineer needs to combine order data from a staging table with customer names from a dimension table and calculate a running total of sales per customer. Which T-SQL feature calculates the running total without collapsing the result set?

A GROUP BY clause with SUM

A window function with SUM ... OVER

A common table expression (CTE)

2.

A team queries the same complex join and aggregation logic from multiple reports and semantic models. What is the primary benefit of implementing this logic as a view instead of repeating the query?

Views store pre-computed results that load faster than running the query directly.

Views define the query once so it can be referenced by name, reducing duplication and simplifying maintenance.

Views enforce foreign key constraints between the tables they reference.

3.

A data engineer builds a stored procedure that accepts year and month parameters to refresh a summary table. The procedure deletes existing rows for that period and inserts freshly aggregated data. Which loading pattern does this procedure implement?

Merge (upsert)

Incremental load

Full refresh of a partition

4.

When loading a fact table from staging data, the INSERT statement joins staging.orders with dim.customer using the filter is_current = 1. What does this filter accomplish?

It restricts the load to only the most recent orders in the staging table.

It links each fact row to the current version of the customer dimension record.

It prevents duplicate rows from being inserted into the fact table.

5.

A data engineer needs to test transformation logic against production warehouse data without risking changes to the original tables. Which Fabric feature supports this scenario with minimal storage overhead?

Create a view over the production table

Create a table clone of the production table

Create a stored procedure that wraps the production table in a transaction

Feedback