Exam Bundle

The collaboration returns approved aggregate results without unrestricted raw extraction would demonstrate the separate action Use a supported Databricks Clean Rooms collaboration with permitted assets and queries, which addresses “The companies need cross-party join analysis with governed restrictions on raw access.” It does not verify the requirement “A proposed workflow is simple one-way distribution with no joint privacy boundary needed” following Use governed Delta Sharing rather than a full Clean Rooms collaboration.

New collaboration queries cannot access the revoked dataset would demonstrate the separate action Remove the authorized asset or access grant from the Clean Room, which addresses “One party must revoke a dataset from an active collaboration.” It does not verify the requirement “A proposed workflow is simple one-way distribution with no joint privacy boundary needed” following Use governed Delta Sharing rather than a full Clean Rooms collaboration.

The prohibited row-level output is not exposed to the participant would demonstrate the separate action Restrict allowable notebook/query execution and output governance in the Clean Room, which addresses “A participant attempts a query that reveals prohibited per-customer attributes.” It does not verify the requirement “A proposed workflow is simple one-way distribution with no joint privacy boundary needed” following Use governed Delta Sharing rather than a full Clean Rooms collaboration.

This action solves a different issue: “The pipeline must reconstruct curated silver tables after a logic bug without losing source history.” Its normal signal is Rebuilt silver counts match the expected corrected output from bronze. Here the requested behavior is “Corrupt JSON records must be preserved for investigation instead of silently dropped,” so that signal is not sufficient.

This action solves a different issue: “An ingestion batch must fail when mandatory transaction identifiers are missing.” Its normal signal is The run fails with an explicit invalid-key error before promoting corrupt rows. Here the requested behavior is “Corrupt JSON records must be preserved for investigation instead of silently dropped,” so that signal is not sufficient.

This action solves a different issue: “An upstream partner re-sends the same file after a network timeout.” Its normal signal is The sink contains one logical record per event even after replay. Here the requested behavior is “Corrupt JSON records must be preserved for investigation instead of silently dropped,” so that signal is not sufficient.

The sink contains one logical record per event even after replay pertains to a different control (Rely on checkpointed ingestion and a deduplicated business key as appropriate) and a different need (An upstream partner re-sends the same file after a network timeout). It cannot establish that Capture malformed payloads in a quarantine or rescued data target met the stated goal.

Rebuilt silver counts match the expected corrected output from bronze pertains to a different control (Preserve immutable bronze input and rerun corrected transformations under controlled refresh) and a different need (The pipeline must reconstruct curated silver tables after a logic bug without losing source history). It cannot establish that Capture malformed payloads in a quarantine or rescued data target met the stated goal.

The run fails with an explicit invalid-key error before promoting corrupt rows would demonstrate the separate action Apply a strict expectation on non-null transaction keys, which addresses “An ingestion batch must fail when mandatory transaction identifiers are missing.” It does not verify the requirement “Corrupt JSON records must be preserved for investigation instead of silently dropped” following Capture malformed payloads in a quarantine or rescued data target.

The alternative Allow the supported schema evolution update and restart as required has a distinct purpose—An optional field appears for the first time and is approved for schema evolution. It would not establish the required outcome (An ingestion batch must fail when mandatory transaction identifiers are missing); the necessary verification concerns The run fails with an explicit invalid-key error before promoting corrupt rows.

The alternative Capture malformed payloads in a quarantine or rescued data target has a distinct purpose—Corrupt JSON records must be preserved for investigation instead of silently dropped. It would not establish the required outcome (An ingestion batch must fail when mandatory transaction identifiers are missing); the necessary verification concerns The run fails with an explicit invalid-key error before promoting corrupt rows.

This action solves a different issue: “The pipeline must reconstruct curated silver tables after a logic bug without losing source history.” Its normal signal is Rebuilt silver counts match the expected corrected output from bronze. Here the requested behavior is “An ingestion batch must fail when mandatory transaction identifiers are missing,” so that signal is not sufficient.

Quarantine retains the offending raw bytes and error reason pertains to a different control (Capture malformed payloads in a quarantine or rescued data target) and a different need (Corrupt JSON records must be preserved for investigation instead of silently dropped). It cannot establish that Apply a strict expectation on non-null transaction keys met the stated goal.

That observation is a valid indicator for “An upstream partner re-sends the same file after a network timeout” when using Rely on checkpointed ingestion and a deduplicated business key as appropriate. The scenario instead tests “An ingestion batch must fail when mandatory transaction identifiers are missing”; the relevant result is The run fails with an explicit invalid-key error before promoting corrupt rows.

That observation is a valid indicator for “The pipeline must reconstruct curated silver tables after a logic bug without losing source history” when using Preserve immutable bronze input and rerun corrected transformations under controlled refresh. The scenario instead tests “An ingestion batch must fail when mandatory transaction identifiers are missing”; the relevant result is The run fails with an explicit invalid-key error before promoting corrupt rows.

The alternative Apply a strict expectation on non-null transaction keys has a distinct purpose—An ingestion batch must fail when mandatory transaction identifiers are missing. It would not establish the required outcome (An optional field appears for the first time and is approved for schema evolution); the necessary verification concerns The target schema includes the approved field with existing rows preserved.

Capture malformed payloads in a quarantine or rescued data target is appropriate when “Corrupt JSON records must be preserved for investigation instead of silently dropped”; it would be verified by Quarantine retains the offending raw bytes and error reason. It does not directly satisfy “An optional field appears for the first time and is approved for schema evolution,” whose relevant acceptance evidence is The target schema includes the approved field with existing rows preserved.

Preserve immutable bronze input and rerun corrected transformations under controlled refresh is appropriate when “The pipeline must reconstruct curated silver tables after a logic bug without losing source history”; it would be verified by Rebuilt silver counts match the expected corrected output from bronze. It does not directly satisfy “An optional field appears for the first time and is approved for schema evolution,” whose relevant acceptance evidence is The target schema includes the approved field with existing rows preserved.

That observation is a valid indicator for “An upstream partner re-sends the same file after a network timeout” when using Rely on checkpointed ingestion and a deduplicated business key as appropriate. The scenario instead tests “An optional field appears for the first time and is approved for schema evolution”; the relevant result is The target schema includes the approved field with existing rows preserved.

The run fails with an explicit invalid-key error before promoting corrupt rows pertains to a different control (Apply a strict expectation on non-null transaction keys) and a different need (An ingestion batch must fail when mandatory transaction identifiers are missing). It cannot establish that Allow the supported schema evolution update and restart as required met the stated goal.

This signal belongs to the adjacent case “The pipeline must reconstruct curated silver tables after a logic bug without losing source history,” whose remedy is Preserve immutable bronze input and rerun corrected transformations under controlled refresh. The observed result for the present requirement is The target schema includes the approved field with existing rows preserved.

Preserve immutable bronze input and rerun corrected transformations under controlled refresh belongs to the separate case “The pipeline must reconstruct curated silver tables after a logic bug without losing source history,” where Rebuilt silver counts match the expected corrected output from bronze would be expected. That differs from the condition in this stem: “An upstream partner re-sends the same file after a network timeout.”

This action solves a different issue: “An ingestion batch must fail when mandatory transaction identifiers are missing.” Its normal signal is The run fails with an explicit invalid-key error before promoting corrupt rows. Here the requested behavior is “An upstream partner re-sends the same file after a network timeout,” so that signal is not sufficient.

Capture malformed payloads in a quarantine or rescued data target is appropriate when “Corrupt JSON records must be preserved for investigation instead of silently dropped”; it would be verified by Quarantine retains the offending raw bytes and error reason. It does not directly satisfy “An upstream partner re-sends the same file after a network timeout,” whose relevant acceptance evidence is The sink contains one logical record per event even after replay.

That observation is a valid indicator for “The pipeline must reconstruct curated silver tables after a logic bug without losing source history” when using Preserve immutable bronze input and rerun corrected transformations under controlled refresh. The scenario instead tests “An upstream partner re-sends the same file after a network timeout”; the relevant result is The sink contains one logical record per event even after replay.

This signal belongs to the adjacent case “An optional field appears for the first time and is approved for schema evolution,” whose remedy is Allow the supported schema evolution update and restart as required. The observed result for the present requirement is The sink contains one logical record per event even after replay.

This signal belongs to the adjacent case “An ingestion batch must fail when mandatory transaction identifiers are missing,” whose remedy is Apply a strict expectation on non-null transaction keys. The observed result for the present requirement is The sink contains one logical record per event even after replay.

This action solves a different issue: “An optional field appears for the first time and is approved for schema evolution.” Its normal signal is The target schema includes the approved field with existing rows preserved. Here the requested behavior is “The pipeline must reconstruct curated silver tables after a logic bug without losing source history,” so that signal is not sufficient.

The alternative Apply a strict expectation on non-null transaction keys has a distinct purpose—An ingestion batch must fail when mandatory transaction identifiers are missing. It would not establish the required outcome (The pipeline must reconstruct curated silver tables after a logic bug without losing source history); the necessary verification concerns Rebuilt silver counts match the expected corrected output from bronze.

The alternative Capture malformed payloads in a quarantine or rescued data target has a distinct purpose—Corrupt JSON records must be preserved for investigation instead of silently dropped. It would not establish the required outcome (The pipeline must reconstruct curated silver tables after a logic bug without losing source history); the necessary verification concerns Rebuilt silver counts match the expected corrected output from bronze.

The run fails with an explicit invalid-key error before promoting corrupt rows pertains to a different control (Apply a strict expectation on non-null transaction keys) and a different need (An ingestion batch must fail when mandatory transaction identifiers are missing). It cannot establish that Preserve immutable bronze input and rerun corrected transformations under controlled refresh met the stated goal.

This signal belongs to the adjacent case “Corrupt JSON records must be preserved for investigation instead of silently dropped,” whose remedy is Capture malformed payloads in a quarantine or rescued data target. The observed result for the present requirement is Rebuilt silver counts match the expected corrected output from bronze.

The sink contains one logical record per event even after replay pertains to a different control (Rely on checkpointed ingestion and a deduplicated business key as appropriate) and a different need (An upstream partner re-sends the same file after a network timeout). It cannot establish that Preserve immutable bronze input and rerun corrected transformations under controlled refresh met the stated goal.