Exam Bundle

Published analytics schemas remain valid while raw inputs evolve pertains to a different control (Keep curated schema contracts at the silver/gold boundary) and a different need (A source schema changes and analysts must be protected from direct raw structure changes). It cannot establish that Reprocess gold from retained validated silver and controlled transformations met the stated goal.

This signal belongs to the adjacent case “Dashboards require stable business-level aggregate tables,” whose remedy is Create gold tables or metric views from validated silver data. The observed result for the present requirement is Rebuilt gold matches reference metrics without re-ingesting source files.

A source event can be retrieved exactly from bronze after a silver reprocessing run pertains to a different control (Persist source-accurate immutable or append-only bronze records with provenance) and a different need (Raw source payloads must remain available for audit and reproducible reprocessing). It cannot establish that Reprocess gold from retained validated silver and controlled transformations met the stated goal.

The alternative Introduce a governed bridge table or appropriate allocation logic has a distinct purpose—A many-to-many relationship between products and promotions causes duplicate fact joins. It would not establish the required outcome (Revenue should be analyzed by date, product and store using a stable relational model); the necessary verification concerns Queries group facts by the correct date, product and store attributes.

This action solves a different issue: “A corrected dimension description should replace prior text with no requirement for history.” Its normal signal is Only the current corrected attribute is returned for the dimension key. Here the requested behavior is “Revenue should be analyzed by date, product and store using a stable relational model,” so that signal is not sufficient.

The alternative Use an unknown/inferred member and reconcile it later has a distinct purpose—A dimension is referenced before its first record arrives from the master system. It would not establish the required outcome (Revenue should be analyzed by date, product and store using a stable relational model); the necessary verification concerns Queries group facts by the correct date, product and store attributes.

Fact revenue totals remain correct when analyzing multiple promotions would demonstrate the separate action Introduce a governed bridge table or appropriate allocation logic, which addresses “A many-to-many relationship between products and promotions causes duplicate fact joins.” It does not verify the requirement “Revenue should be analyzed by date, product and store using a stable relational model” following Build a sales fact table linked to conformed dimensions by surrogate keys.

This signal belongs to the adjacent case “A dimension is referenced before its first record arrives from the master system,” whose remedy is Use an unknown/inferred member and reconcile it later. The observed result for the present requirement is Queries group facts by the correct date, product and store attributes.

That observation is a valid indicator for “A corrected dimension description should replace prior text with no requirement for history” when using Use SCD Type 1 overwrite semantics for that attribute. The scenario instead tests “Revenue should be analyzed by date, product and store using a stable relational model”; the relevant result is Queries group facts by the correct date, product and store attributes.

Introduce a governed bridge table or appropriate allocation logic belongs to the separate case “A many-to-many relationship between products and promotions causes duplicate fact joins,” where Fact revenue totals remain correct when analyzing multiple promotions would be expected. That differs from the condition in this stem: “A descriptive customer attribute changes and historical analysis must preserve earlier values.”

Build a sales fact table linked to conformed dimensions by surrogate keys belongs to the separate case “Revenue should be analyzed by date, product and store using a stable relational model,” where Queries group facts by the correct date, product and store attributes would be expected. That differs from the condition in this stem: “A descriptive customer attribute changes and historical analysis must preserve earlier values.”

Use an unknown/inferred member and reconcile it later belongs to the separate case “A dimension is referenced before its first record arrives from the master system,” where Fact foreign keys remain valid and eventually resolve to the intended member would be expected. That differs from the condition in this stem: “A descriptive customer attribute changes and historical analysis must preserve earlier values.”

Only the current corrected attribute is returned for the dimension key would demonstrate the separate action Use SCD Type 1 overwrite semantics for that attribute, which addresses “A corrected dimension description should replace prior text with no requirement for history.” It does not verify the requirement “A descriptive customer attribute changes and historical analysis must preserve earlier values” following Use an SCD Type 2 customer dimension with validity windows.

This signal belongs to the adjacent case “A dimension is referenced before its first record arrives from the master system,” whose remedy is Use an unknown/inferred member and reconcile it later. The observed result for the present requirement is Historical facts resolve to the customer version active at transaction time.

That observation is a valid indicator for “A many-to-many relationship between products and promotions causes duplicate fact joins” when using Introduce a governed bridge table or appropriate allocation logic. The scenario instead tests “A descriptive customer attribute changes and historical analysis must preserve earlier values”; the relevant result is Historical facts resolve to the customer version active at transaction time.

Use an unknown/inferred member and reconcile it later belongs to the separate case “A dimension is referenced before its first record arrives from the master system,” where Fact foreign keys remain valid and eventually resolve to the intended member would be expected. That differs from the condition in this stem: “A corrected dimension description should replace prior text with no requirement for history.”

This action solves a different issue: “A many-to-many relationship between products and promotions causes duplicate fact joins.” Its normal signal is Fact revenue totals remain correct when analyzing multiple promotions. Here the requested behavior is “A corrected dimension description should replace prior text with no requirement for history,” so that signal is not sufficient.

This action solves a different issue: “Revenue should be analyzed by date, product and store using a stable relational model.” Its normal signal is Queries group facts by the correct date, product and store attributes. Here the requested behavior is “A corrected dimension description should replace prior text with no requirement for history,” so that signal is not sufficient.

Queries group facts by the correct date, product and store attributes would demonstrate the separate action Build a sales fact table linked to conformed dimensions by surrogate keys, which addresses “Revenue should be analyzed by date, product and store using a stable relational model.” It does not verify the requirement “A corrected dimension description should replace prior text with no requirement for history” following Use SCD Type 1 overwrite semantics for that attribute.

This signal belongs to the adjacent case “A dimension is referenced before its first record arrives from the master system,” whose remedy is Use an unknown/inferred member and reconcile it later. The observed result for the present requirement is Only the current corrected attribute is returned for the dimension key.

This signal belongs to the adjacent case “A descriptive customer attribute changes and historical analysis must preserve earlier values,” whose remedy is Use an SCD Type 2 customer dimension with validity windows. The observed result for the present requirement is Only the current corrected attribute is returned for the dimension key.

Use an SCD Type 2 customer dimension with validity windows is appropriate when “A descriptive customer attribute changes and historical analysis must preserve earlier values”; it would be verified by Historical facts resolve to the customer version active at transaction time. It does not directly satisfy “A dimension is referenced before its first record arrives from the master system,” whose relevant acceptance evidence is Fact foreign keys remain valid and eventually resolve to the intended member.

Build a sales fact table linked to conformed dimensions by surrogate keys is appropriate when “Revenue should be analyzed by date, product and store using a stable relational model”; it would be verified by Queries group facts by the correct date, product and store attributes. It does not directly satisfy “A dimension is referenced before its first record arrives from the master system,” whose relevant acceptance evidence is Fact foreign keys remain valid and eventually resolve to the intended member.

The alternative Use SCD Type 1 overwrite semantics for that attribute has a distinct purpose—A corrected dimension description should replace prior text with no requirement for history. It would not establish the required outcome (A dimension is referenced before its first record arrives from the master system); the necessary verification concerns Fact foreign keys remain valid and eventually resolve to the intended member.

Historical facts resolve to the customer version active at transaction time would demonstrate the separate action Use an SCD Type 2 customer dimension with validity windows, which addresses “A descriptive customer attribute changes and historical analysis must preserve earlier values.” It does not verify the requirement “A dimension is referenced before its first record arrives from the master system” following Use an unknown/inferred member and reconcile it later.

This signal belongs to the adjacent case “Revenue should be analyzed by date, product and store using a stable relational model,” whose remedy is Build a sales fact table linked to conformed dimensions by surrogate keys. The observed result for the present requirement is Fact foreign keys remain valid and eventually resolve to the intended member.

Only the current corrected attribute is returned for the dimension key pertains to a different control (Use SCD Type 1 overwrite semantics for that attribute) and a different need (A corrected dimension description should replace prior text with no requirement for history). It cannot establish that Use an unknown/inferred member and reconcile it later met the stated goal.

The alternative Use an unknown/inferred member and reconcile it later has a distinct purpose—A dimension is referenced before its first record arrives from the master system. It would not establish the required outcome (A many-to-many relationship between products and promotions causes duplicate fact joins); the necessary verification concerns Fact revenue totals remain correct when analyzing multiple promotions.

The alternative Use an SCD Type 2 customer dimension with validity windows has a distinct purpose—A descriptive customer attribute changes and historical analysis must preserve earlier values. It would not establish the required outcome (A many-to-many relationship between products and promotions causes duplicate fact joins); the necessary verification concerns Fact revenue totals remain correct when analyzing multiple promotions.

Build a sales fact table linked to conformed dimensions by surrogate keys is appropriate when “Revenue should be analyzed by date, product and store using a stable relational model”; it would be verified by Queries group facts by the correct date, product and store attributes. It does not directly satisfy “A many-to-many relationship between products and promotions causes duplicate fact joins,” whose relevant acceptance evidence is Fact revenue totals remain correct when analyzing multiple promotions.

This signal belongs to the adjacent case “A dimension is referenced before its first record arrives from the master system,” whose remedy is Use an unknown/inferred member and reconcile it later. The observed result for the present requirement is Fact revenue totals remain correct when analyzing multiple promotions.

That observation is a valid indicator for “A corrected dimension description should replace prior text with no requirement for history” when using Use SCD Type 1 overwrite semantics for that attribute. The scenario instead tests “A many-to-many relationship between products and promotions causes duplicate fact joins”; the relevant result is Fact revenue totals remain correct when analyzing multiple promotions.

Queries group facts by the correct date, product and store attributes would demonstrate the separate action Build a sales fact table linked to conformed dimensions by surrogate keys, which addresses “Revenue should be analyzed by date, product and store using a stable relational model.” It does not verify the requirement “A many-to-many relationship between products and promotions causes duplicate fact joins” following Introduce a governed bridge table or appropriate allocation logic.