Exam Bundle

This signal belongs to the adjacent case “A personal-data column must not be sent to an unapproved inference endpoint,” whose remedy is Mask or exclude protected fields and use approved governed model access. The observed result for the present requirement is Regression tests detect a change in label accuracy or format compliance.

This signal belongs to the adjacent case “A model response is free text but downstream processing requires a strict schema,” whose remedy is Validate and parse the model response against an explicit output contract. The observed result for the present requirement is Regression tests detect a change in label accuracy or format compliance.

That observation is a valid indicator for “AI results should not be recomputed for every dashboard query” when using Materialize validated model outputs in a governed table on a controlled schedule. The scenario instead tests “AI-generated labels are inconsistent and need acceptance evidence”; the relevant result is Regression tests detect a change in label accuracy or format compliance.

Use `ON VIOLATION FAIL UPDATE` or equivalent strict expectation belongs to the separate case “A critical primary-key validation must stop the entire pipeline on failure,” where The update stops with the named expectation failure would be expected. That differs from the condition in this stem: “All failing records should be counted but retained in a diagnostic bronze target.”

Write failed rows to a dedicated quarantine table with failure metadata is appropriate when “Auditors require the actual rejected input rows and reason codes”; it would be verified by Quarantine identifies each invalid source record and violated rule. It does not directly satisfy “All failing records should be counted but retained in a diagnostic bronze target,” whose relevant acceptance evidence is The expectation violation count rises while raw records stay available.

This action solves a different issue: “An expectation is misconfigured to reject legal nullable optional fields.” Its normal signal is Valid optional-null records remain while truly invalid required fields are rejected. Here the requested behavior is “All failing records should be counted but retained in a diagnostic bronze target,” so that signal is not sufficient.

The silver output excludes the rejected rows while pipeline proceeds pertains to a different control (Use `ON VIOLATION DROP ROW` or an equivalent supported expectation setting) and a different need (A valid silver table should omit rows that fail a noncritical quality predicate). It cannot establish that Use a warning expectation and retain the raw rows met the stated goal.

That observation is a valid indicator for “An expectation is misconfigured to reject legal nullable optional fields” when using Restrict the predicate to required columns per the data contract. The scenario instead tests “All failing records should be counted but retained in a diagnostic bronze target”; the relevant result is The expectation violation count rises while raw records stay available.

This signal belongs to the adjacent case “A critical primary-key validation must stop the entire pipeline on failure,” whose remedy is Use `ON VIOLATION FAIL UPDATE` or equivalent strict expectation. The observed result for the present requirement is The expectation violation count rises while raw records stay available.

The alternative Use a warning expectation and retain the raw rows has a distinct purpose—All failing records should be counted but retained in a diagnostic bronze target. It would not establish the required outcome (A valid silver table should omit rows that fail a noncritical quality predicate); the necessary verification concerns The silver output excludes the rejected rows while pipeline proceeds.

The alternative Write failed rows to a dedicated quarantine table with failure metadata has a distinct purpose—Auditors require the actual rejected input rows and reason codes. It would not establish the required outcome (A valid silver table should omit rows that fail a noncritical quality predicate); the necessary verification concerns The silver output excludes the rejected rows while pipeline proceeds.

This action solves a different issue: “An expectation is misconfigured to reject legal nullable optional fields.” Its normal signal is Valid optional-null records remain while truly invalid required fields are rejected. Here the requested behavior is “A valid silver table should omit rows that fail a noncritical quality predicate,” so that signal is not sufficient.

The expectation violation count rises while raw records stay available would demonstrate the separate action Use a warning expectation and retain the raw rows, which addresses “All failing records should be counted but retained in a diagnostic bronze target.” It does not verify the requirement “A valid silver table should omit rows that fail a noncritical quality predicate” following Use `ON VIOLATION DROP ROW` or an equivalent supported expectation setting.

This signal belongs to the adjacent case “A critical primary-key validation must stop the entire pipeline on failure,” whose remedy is Use `ON VIOLATION FAIL UPDATE` or equivalent strict expectation. The observed result for the present requirement is The silver output excludes the rejected rows while pipeline proceeds.

This signal belongs to the adjacent case “An expectation is misconfigured to reject legal nullable optional fields,” whose remedy is Restrict the predicate to required columns per the data contract. The observed result for the present requirement is The silver output excludes the rejected rows while pipeline proceeds.

The alternative Use a warning expectation and retain the raw rows has a distinct purpose—All failing records should be counted but retained in a diagnostic bronze target. It would not establish the required outcome (A critical primary-key validation must stop the entire pipeline on failure); the necessary verification concerns The update stops with the named expectation failure.

The alternative Write failed rows to a dedicated quarantine table with failure metadata has a distinct purpose—Auditors require the actual rejected input rows and reason codes. It would not establish the required outcome (A critical primary-key validation must stop the entire pipeline on failure); the necessary verification concerns The update stops with the named expectation failure.

Restrict the predicate to required columns per the data contract belongs to the separate case “An expectation is misconfigured to reject legal nullable optional fields,” where Valid optional-null records remain while truly invalid required fields are rejected would be expected. That differs from the condition in this stem: “A critical primary-key validation must stop the entire pipeline on failure.”

That observation is a valid indicator for “All failing records should be counted but retained in a diagnostic bronze target” when using Use a warning expectation and retain the raw rows. The scenario instead tests “A critical primary-key validation must stop the entire pipeline on failure”; the relevant result is The update stops with the named expectation failure.

Valid optional-null records remain while truly invalid required fields are rejected would demonstrate the separate action Restrict the predicate to required columns per the data contract, which addresses “An expectation is misconfigured to reject legal nullable optional fields.” It does not verify the requirement “A critical primary-key validation must stop the entire pipeline on failure” following Use `ON VIOLATION FAIL UPDATE` or equivalent strict expectation.

This signal belongs to the adjacent case “Auditors require the actual rejected input rows and reason codes,” whose remedy is Write failed rows to a dedicated quarantine table with failure metadata. The observed result for the present requirement is The update stops with the named expectation failure.

This action solves a different issue: “All failing records should be counted but retained in a diagnostic bronze target.” Its normal signal is The expectation violation count rises while raw records stay available. Here the requested behavior is “Auditors require the actual rejected input rows and reason codes,” so that signal is not sufficient.

Use `ON VIOLATION DROP ROW` or an equivalent supported expectation setting is appropriate when “A valid silver table should omit rows that fail a noncritical quality predicate”; it would be verified by The silver output excludes the rejected rows while pipeline proceeds. It does not directly satisfy “Auditors require the actual rejected input rows and reason codes,” whose relevant acceptance evidence is Quarantine identifies each invalid source record and violated rule.

Restrict the predicate to required columns per the data contract belongs to the separate case “An expectation is misconfigured to reject legal nullable optional fields,” where Valid optional-null records remain while truly invalid required fields are rejected would be expected. That differs from the condition in this stem: “Auditors require the actual rejected input rows and reason codes.”

The update stops with the named expectation failure would demonstrate the separate action Use `ON VIOLATION FAIL UPDATE` or equivalent strict expectation, which addresses “A critical primary-key validation must stop the entire pipeline on failure.” It does not verify the requirement “Auditors require the actual rejected input rows and reason codes” following Write failed rows to a dedicated quarantine table with failure metadata.

Valid optional-null records remain while truly invalid required fields are rejected pertains to a different control (Restrict the predicate to required columns per the data contract) and a different need (An expectation is misconfigured to reject legal nullable optional fields). It cannot establish that Write failed rows to a dedicated quarantine table with failure metadata met the stated goal.

This signal belongs to the adjacent case “A valid silver table should omit rows that fail a noncritical quality predicate,” whose remedy is Use `ON VIOLATION DROP ROW` or an equivalent supported expectation setting. The observed result for the present requirement is Quarantine identifies each invalid source record and violated rule.

Use `ON VIOLATION FAIL UPDATE` or equivalent strict expectation belongs to the separate case “A critical primary-key validation must stop the entire pipeline on failure,” where The update stops with the named expectation failure would be expected. That differs from the condition in this stem: “An expectation is misconfigured to reject legal nullable optional fields.”

Use a warning expectation and retain the raw rows is appropriate when “All failing records should be counted but retained in a diagnostic bronze target”; it would be verified by The expectation violation count rises while raw records stay available. It does not directly satisfy “An expectation is misconfigured to reject legal nullable optional fields,” whose relevant acceptance evidence is Valid optional-null records remain while truly invalid required fields are rejected.

Write failed rows to a dedicated quarantine table with failure metadata belongs to the separate case “Auditors require the actual rejected input rows and reason codes,” where Quarantine identifies each invalid source record and violated rule would be expected. That differs from the condition in this stem: “An expectation is misconfigured to reject legal nullable optional fields.”

The silver output excludes the rejected rows while pipeline proceeds would demonstrate the separate action Use `ON VIOLATION DROP ROW` or an equivalent supported expectation setting, which addresses “A valid silver table should omit rows that fail a noncritical quality predicate.” It does not verify the requirement “An expectation is misconfigured to reject legal nullable optional fields” following Restrict the predicate to required columns per the data contract.

The expectation violation count rises while raw records stay available would demonstrate the separate action Use a warning expectation and retain the raw rows, which addresses “All failing records should be counted but retained in a diagnostic bronze target.” It does not verify the requirement “An expectation is misconfigured to reject legal nullable optional fields” following Restrict the predicate to required columns per the data contract.

The update stops with the named expectation failure pertains to a different control (Use `ON VIOLATION FAIL UPDATE` or equivalent strict expectation) and a different need (A critical primary-key validation must stop the entire pipeline on failure). It cannot establish that Restrict the predicate to required columns per the data contract met the stated goal.