Question 700 A dimension stores rate records with valid_from and valid_to dates. An order created in March must use the rate valid on the order date, not the newest rate today. Which condition should the join include? Business key equality alone Rate value greater than zero only Order date after the newest current date only Order date within the dimension validity interval, in addition to the business key Explanation Correct Answer D:A temporal join selects the historically applicable rate for the transaction. Incorrect Answers A:A key-only join can match multiple historical rate rows and inflate orders. B:A positive rate is not evidence that it was effective on the given order date. C:This selects the wrong rate for historical transactions before the latest change.
Question 701 A fact table contains 5 transactions, and an engineer groups by country across two distinct countries. Assuming no null filters, how many rows does the grouped result contain? 5 10 1 2 Explanation Correct Answer D:GroupBy produces one aggregate row per distinct grouping key, hence two country rows. Incorrect Answers A:Five is the source transaction row count, not the number of grouped countries. B:Grouping does not multiply source rows by the number of country values. C:One aggregate would occur if no grouping key were used, not when two countries are present.
Question 702 A dataset has category A with customer IDs [1,1,2] and category B with IDs [2,3]. A groupBy(category).countDistinct(customer_id) is run. What are the two category counts? A=2, B=2 A=3, B=2 A=3, B=3 A=2, B=1 Explanation Correct Answer A:A contains IDs 1 and 2; B contains 2 and 3. Distinct counts are calculated independently per group. Incorrect Answers B:Three counts rows in category A and includes its repeated customer ID twice. C:Neither group has three distinct customers under the listed values. D:Category B includes both 2 and 3, not just one distinct customer.
Question 703 A silver pipeline rejects ten malformed invoices and records them in a quarantine table. The gold team asks why totals do not match raw ingestion counts. What should the reconciliation report show? Claim all raw invoices appeared in gold regardless of rejects Accepted, rejected, and total source counts with explicit rejection reasons Change the report label without adjusting metrics Delete quarantine records to remove the discrepancy Explanation Correct Answer B:A controlled reconciliation can explain how raw input splits into publishable data and quarantined exceptions. Incorrect Answers A:This misstates the accepted population and conceals the ten invalid records. C:A renamed metric does not explain the source-to-silver row-level reconciliation. D:Destroying exception evidence makes discrepancies harder to audit and remediate.
Question 704 A gold report shows region revenue A=30 and B=70, total=100. What share of total revenue comes from A? 43% 100% 30% 70% Explanation Correct Answer C:Region A contributes 30 of the overall 100 revenue units, a share of 0.30. Incorrect Answers A:Dividing 30 by 70 compares regions but is not A's share of the overall total. B:The full total includes A and B, so A alone cannot account for every unit. D:Seventy percent is B's share, not region A's contribution.
Question 705 A dashboard combines team A average latency 100ms over 10 requests and team B average latency 200ms over 90 requests. What is the correct overall average latency? 150 ms 100 ms 200 ms 190 ms Explanation Correct Answer D:The weighted mean is (100×10 + 200×90)/100 = 190 milliseconds. Incorrect Answers A:A simple mean of the two team averages ignores their unequal request counts. B:This uses only team A and ignores the ninety requests from team B. C:This uses only team B and ignores ten valid requests from team A.
Question 706 A source has transaction rows with customer_id NULL. A downstream inner join to a customer table omits these rows. Which silver-stage action best preserves a trace of the omission? Delete the source table after the join Substitute the same fake customer ID for every missing key Explicitly identify and quarantine missing-key transactions before the join Assume the inner join preserves every source row Explanation Correct Answer C:A pre-join key-quality check records why some transactions cannot be matched. Incorrect Answers A:Deleting source data removes evidence needed to diagnose the missing transactions. B:A shared invented key causes unrelated transactions to merge under a false identity. D:An inner join discards source rows without a matching customer key.
Question 707 A silver engineer designs a daily inventory snapshot at warehouse_id, product_id, snapshot_date grain. What uniqueness check should be used before the gold build? Ensure the input filename is unique across all rows Ensure no duplicate rows for the composite warehouse-product-date key Ensure stock_quantity is unique across all products Ensure product_id is globally unique across all dates and warehouses Explanation Correct Answer B:The three fields together define exactly one business snapshot under the agreed grain. Incorrect Answers A:Many snapshot rows can correctly originate in one file; filename is not the business key. C:Different products can have equal quantities without violating snapshot identity. D:The same product legitimately occurs at multiple warehouses and on different dates.
Question 708 A gold table must sort months chronologically, but the source month labels are "1", "10", "2". What should the engineer do? Aggregate by source file extension Store or derive a numeric/date month key and sort by that value Sort month strings ascending and assume numeric order Sort by the original row arrival order Explanation Correct Answer B:A typed month ordering produces 1,2,10 rather than lexicographic 1,10,2. Incorrect Answers A:File extensions do not represent temporal ordering of month values. C:Text ordering places "10" before "2", contrary to chronological order. D:Ingestion order does not encode calendar sequence reliably.
Question 709 A key with extremely frequent values makes one join partition much larger than others. A proposed repartition by the same unmodified key does not help. Why? Join logic stops comparing keys after repartition Repartition always ensures every key is split evenly across workers All records sharing the hot key can still land in the same partition Repartition automatically removes duplicate business events Explanation Correct Answer C:Hash partitioning on the unchanged skewed key retains concentration for that hot value. Incorrect Answers A:Repartition changes data movement and placement, not the equality conditions of the join. B:A key-based partitioner normally keeps identical key values together. D:Physical distribution does not implement record-level deduplication.
Question 710 A gold view is created over a sensitive silver table. A requester has visibility into the view definition but no usable permission path to underlying data. What should be reviewed? The applicable view access and underlying authorization semantics for that execution context Review only the view’s SQL definition without checking its ownership and applicable source authorization Assume the view definition grants all underlying source privileges automatically Only the number of task retries Explanation Correct Answer A:A saved view name alone does not guarantee every principal can query protected data. Incorrect Answers B:A visible view definition is not proof that an executing principal has the necessary permissions in the current view security context. C:View visibility alone cannot establish the effective authorization path needed in the applicable execution context. D:Retries cannot fix a predictable missing permission for the data source.