Exam Bundle

The pipeline authenticates as the restricted service identity, not admin would demonstrate the separate action Create a dedicated least-privilege replication identity for the UC connection, which addresses “An administrator uses root database credentials inside the production ingestion pipeline.” It does not verify the requirement “An updated source table is missing from a configured publication” following Add that authorized table to the CDC publication and refresh gateway metadata.

This signal belongs to the adjacent case “The database ingestion user lacks permissions to read the required tables,” whose remedy is Grant only the documented replication and source-read privileges. The observed result for the present requirement is Changes for the newly included table flow into its intended target.

Replication slot lag and WAL retention stay below the approved threshold would demonstrate the separate action Monitor replication lag and control WAL retention and slot lifecycle, which addresses “An idle PostgreSQL logical replication slot retains WAL indefinitely and threatens source disk.” It does not verify the requirement “An updated source table is missing from a configured publication” following Add that authorized table to the CDC publication and refresh gateway metadata.

Capture file path, ingest timestamp and source system metadata in bronze is appropriate when “A vendor requires durable raw provenance for every file used to build a curated table”; it would be verified by Each curated record can be traced to its raw source record. It does not directly satisfy “A new bronze pipeline requires transactionally consistent append-only event storage,” whose relevant acceptance evidence is Bronze consumers read a complete consistent version after each batch.

The alternative Normalize the key field during bronze-to-silver ingestion has a distinct purpose—Several operational feeds use differing column names for the same business key. It would not establish the required outcome (A new bronze pipeline requires transactionally consistent append-only event storage); the necessary verification concerns Bronze consumers read a complete consistent version after each batch.

Use a supported Iceberg reader or managed interoperability approach belongs to the separate case “Data producers supply Iceberg snapshots that Databricks must consume under supported interoperability,” where The reader returns the expected committed Iceberg snapshot rows would be expected. That differs from the condition in this stem: “A new bronze pipeline requires transactionally consistent append-only event storage.”

Each curated record can be traced to its raw source record pertains to a different control (Capture file path, ingest timestamp and source system metadata in bronze) and a different need (A vendor requires durable raw provenance for every file used to build a curated table). It cannot establish that Write ingested records into a Delta table with committed transactions met the stated goal.

The reader returns the expected committed Iceberg snapshot rows would demonstrate the separate action Use a supported Iceberg reader or managed interoperability approach, which addresses “Data producers supply Iceberg snapshots that Databricks must consume under supported interoperability.” It does not verify the requirement “A new bronze pipeline requires transactionally consistent append-only event storage” following Write ingested records into a Delta table with committed transactions.

The consumer sees inserts, updates and deletes after the selected version would demonstrate the separate action Enable and consume Delta Change Data Feed with supported version tracking, which addresses “A downstream consumer needs fresh changes from an existing Delta table instead of complete copies.” It does not verify the requirement “A new bronze pipeline requires transactionally consistent append-only event storage” following Write ingested records into a Delta table with committed transactions.

Enable and consume Delta Change Data Feed with supported version tracking is appropriate when “A downstream consumer needs fresh changes from an existing Delta table instead of complete copies”; it would be verified by The consumer sees inserts, updates and deletes after the selected version. It does not directly satisfy “Data producers supply Iceberg snapshots that Databricks must consume under supported interoperability,” whose relevant acceptance evidence is The reader returns the expected committed Iceberg snapshot rows.

The alternative Capture file path, ingest timestamp and source system metadata in bronze has a distinct purpose—A vendor requires durable raw provenance for every file used to build a curated table. It would not establish the required outcome (Data producers supply Iceberg snapshots that Databricks must consume under supported interoperability); the necessary verification concerns The reader returns the expected committed Iceberg snapshot rows.

The alternative Normalize the key field during bronze-to-silver ingestion has a distinct purpose—Several operational feeds use differing column names for the same business key. It would not establish the required outcome (Data producers supply Iceberg snapshots that Databricks must consume under supported interoperability); the necessary verification concerns The reader returns the expected committed Iceberg snapshot rows.

That observation is a valid indicator for “Several operational feeds use differing column names for the same business key” when using Normalize the key field during bronze-to-silver ingestion. The scenario instead tests “Data producers supply Iceberg snapshots that Databricks must consume under supported interoperability”; the relevant result is The reader returns the expected committed Iceberg snapshot rows.

Each curated record can be traced to its raw source record would demonstrate the separate action Capture file path, ingest timestamp and source system metadata in bronze, which addresses “A vendor requires durable raw provenance for every file used to build a curated table.” It does not verify the requirement “Data producers supply Iceberg snapshots that Databricks must consume under supported interoperability” following Use a supported Iceberg reader or managed interoperability approach.

This signal belongs to the adjacent case “A new bronze pipeline requires transactionally consistent append-only event storage,” whose remedy is Write ingested records into a Delta table with committed transactions. The observed result for the present requirement is The reader returns the expected committed Iceberg snapshot rows.

The alternative Normalize the key field during bronze-to-silver ingestion has a distinct purpose—Several operational feeds use differing column names for the same business key. It would not establish the required outcome (A downstream consumer needs fresh changes from an existing Delta table instead of complete copies); the necessary verification concerns The consumer sees inserts, updates and deletes after the selected version.

Capture file path, ingest timestamp and source system metadata in bronze belongs to the separate case “A vendor requires durable raw provenance for every file used to build a curated table,” where Each curated record can be traced to its raw source record would be expected. That differs from the condition in this stem: “A downstream consumer needs fresh changes from an existing Delta table instead of complete copies.”

Use a supported Iceberg reader or managed interoperability approach belongs to the separate case “Data producers supply Iceberg snapshots that Databricks must consume under supported interoperability,” where The reader returns the expected committed Iceberg snapshot rows would be expected. That differs from the condition in this stem: “A downstream consumer needs fresh changes from an existing Delta table instead of complete copies.”

This signal belongs to the adjacent case “Several operational feeds use differing column names for the same business key,” whose remedy is Normalize the key field during bronze-to-silver ingestion. The observed result for the present requirement is The consumer sees inserts, updates and deletes after the selected version.

The reader returns the expected committed Iceberg snapshot rows would demonstrate the separate action Use a supported Iceberg reader or managed interoperability approach, which addresses “Data producers supply Iceberg snapshots that Databricks must consume under supported interoperability.” It does not verify the requirement “A downstream consumer needs fresh changes from an existing Delta table instead of complete copies” following Enable and consume Delta Change Data Feed with supported version tracking.

That observation is a valid indicator for “A vendor requires durable raw provenance for every file used to build a curated table” when using Capture file path, ingest timestamp and source system metadata in bronze. The scenario instead tests “A downstream consumer needs fresh changes from an existing Delta table instead of complete copies”; the relevant result is The consumer sees inserts, updates and deletes after the selected version.

Enable and consume Delta Change Data Feed with supported version tracking is appropriate when “A downstream consumer needs fresh changes from an existing Delta table instead of complete copies”; it would be verified by The consumer sees inserts, updates and deletes after the selected version. It does not directly satisfy “Several operational feeds use differing column names for the same business key,” whose relevant acceptance evidence is The silver target contains one consistent key type and name.

This action solves a different issue: “A new bronze pipeline requires transactionally consistent append-only event storage.” Its normal signal is Bronze consumers read a complete consistent version after each batch. Here the requested behavior is “Several operational feeds use differing column names for the same business key,” so that signal is not sufficient.

This action solves a different issue: “A vendor requires durable raw provenance for every file used to build a curated table.” Its normal signal is Each curated record can be traced to its raw source record. Here the requested behavior is “Several operational feeds use differing column names for the same business key,” so that signal is not sufficient.

That observation is a valid indicator for “A downstream consumer needs fresh changes from an existing Delta table instead of complete copies” when using Enable and consume Delta Change Data Feed with supported version tracking. The scenario instead tests “Several operational feeds use differing column names for the same business key”; the relevant result is The silver target contains one consistent key type and name.

Each curated record can be traced to its raw source record would demonstrate the separate action Capture file path, ingest timestamp and source system metadata in bronze, which addresses “A vendor requires durable raw provenance for every file used to build a curated table.” It does not verify the requirement “Several operational feeds use differing column names for the same business key” following Normalize the key field during bronze-to-silver ingestion.

This signal belongs to the adjacent case “Data producers supply Iceberg snapshots that Databricks must consume under supported interoperability,” whose remedy is Use a supported Iceberg reader or managed interoperability approach. The observed result for the present requirement is The silver target contains one consistent key type and name.

Use a supported Iceberg reader or managed interoperability approach belongs to the separate case “Data producers supply Iceberg snapshots that Databricks must consume under supported interoperability,” where The reader returns the expected committed Iceberg snapshot rows would be expected. That differs from the condition in this stem: “A vendor requires durable raw provenance for every file used to build a curated table.”

Normalize the key field during bronze-to-silver ingestion is appropriate when “Several operational feeds use differing column names for the same business key”; it would be verified by The silver target contains one consistent key type and name. It does not directly satisfy “A vendor requires durable raw provenance for every file used to build a curated table,” whose relevant acceptance evidence is Each curated record can be traced to its raw source record.

Enable and consume Delta Change Data Feed with supported version tracking belongs to the separate case “A downstream consumer needs fresh changes from an existing Delta table instead of complete copies,” where The consumer sees inserts, updates and deletes after the selected version would be expected. That differs from the condition in this stem: “A vendor requires durable raw provenance for every file used to build a curated table.”

That observation is a valid indicator for “A downstream consumer needs fresh changes from an existing Delta table instead of complete copies” when using Enable and consume Delta Change Data Feed with supported version tracking. The scenario instead tests “A vendor requires durable raw provenance for every file used to build a curated table”; the relevant result is Each curated record can be traced to its raw source record.

The silver target contains one consistent key type and name would demonstrate the separate action Normalize the key field during bronze-to-silver ingestion, which addresses “Several operational feeds use differing column names for the same business key.” It does not verify the requirement “A vendor requires durable raw provenance for every file used to build a curated table” following Capture file path, ingest timestamp and source system metadata in bronze.

That observation is a valid indicator for “A new bronze pipeline requires transactionally consistent append-only event storage” when using Write ingested records into a Delta table with committed transactions. The scenario instead tests “A vendor requires durable raw provenance for every file used to build a curated table”; the relevant result is Each curated record can be traced to its raw source record.