SnowPro Core COF-C03 Practice Test 96
Question 950
A deployment team has configured this Snowflake capability: Inspect COPY_HISTORY or supported load history and detailed COPY results. After promoting the configuration to the intended environment, auditors want observable evidence rather than a successful Save click. Which test outcome most convincingly verifies the capability is functioning as intended?
- A published Kafka test event is visible in Snowflake within the target freshness window.
- Connection logs show policy rejection for the blocked origin and SQL execution after approved access.
- The target file's recorded status, error and load time match the failed job trace.
- An update to the downstream consumer causes the intermediate table to refresh in dependency order.
Explanation
Correct Answer
C:Load history provides observable file-level execution and error outcomes for ingestion investigations. The stated acceptance evidence is: The target file's recorded status, error and load time match the failed job trace. Operational verification is stronger than an assumption based on saving a configuration.
Incorrect Answers
A:The connector can route event rows to Snowflake's streaming ingestion instead of warehouse COPY batches. That explains the role of “A published Kafka test event is visible in Snowflake within the target freshness window”, but the present requirement calls for “The target file's recorded status, error and load time match the failed job trace”: Load history provides observable file-level execution and error outcomes for ingestion investigations.
B:Network restrictions can reject a connection before a query reaches Snowflake's optimizer. That explains the role of “Connection logs show policy rejection for the blocked origin and SQL execution after approved access”, but the present requirement calls for “The target file's recorded status, error and load time match the failed job trace”: Load history provides observable file-level execution and error outcomes for ingestion investigations.
D:DOWNSTREAM delegates refresh scheduling to dependent tables instead of maintaining an independent timer. That explains the role of “An update to the downstream consumer causes the intermediate table to refresh in dependency order”, but the present requirement calls for “The target file's recorded status, error and load time match the failed job trace”: Load history provides observable file-level execution and error outcomes for ingestion investigations.
Question 951
A deployment team has configured this Snowflake capability: Control for result cache, cold starts, query text and comparable input data before comparing warehouse sizes. After promoting the configuration to the intended environment, auditors want observable evidence rather than a successful Save click. Which test outcome most convincingly verifies the capability is functioning as intended?
- Repeated equivalent uncached runs show stable latency and credit consumption for both sizes.
- The revised query returns expected unique business keys and reconciled totals.
- Representative JSON records yield the expected numeric and date values.
- Usage history shows no long idle runtime and morning queries resume automatically.
Explanation
Correct Answer
A:A fair warehouse performance test must separate cache effects from the resource configuration change. The stated acceptance evidence is: Repeated equivalent uncached runs show stable latency and credit consumption for both sizes. Operational verification is stronger than an assumption based on saving a configuration.
Incorrect Answers
B:Joining a one-to-many table without accounting for cardinality can multiply logical fact values. That explains the role of “The revised query returns expected unique business keys and reconciled totals”, but the present requirement calls for “Repeated equivalent uncached runs show stable latency and credit consumption for both sizes”: A fair warehouse performance test must separate cache effects from the resource configuration change.
C:Typed extraction converts flexible input data into stable columns for reporting and validation. That explains the role of “Representative JSON records yield the expected numeric and date values”, but the present requirement calls for “Repeated equivalent uncached runs show stable latency and credit consumption for both sizes”: A fair warehouse performance test must separate cache effects from the resource configuration change.
D:Suspension releases active compute after an idle period rather than relying on staff to stop it. That explains the role of “Usage history shows no long idle runtime and morning queries resume automatically”, but the present requirement calls for “Repeated equivalent uncached runs show stable latency and credit consumption for both sizes”: A fair warehouse performance test must separate cache effects from the resource configuration change.
Question 952
A deployment team has configured this Snowflake capability: Verify share grants and the required database and schema object privileges in the provider account. After promoting the configuration to the intended environment, auditors want observable evidence rather than a successful Save click. Which test outcome most convincingly verifies the capability is functioning as intended?
- A consumer refreshes the imported share and can SELECT the newly granted secure view.
- Source table row counts remain stable while the clone reflects only the test insert.
- A disaster-recovery drill can access the replicated secondary account after loss of the primary region.
- Object metadata reflects transient type and the retention policy matches the accepted operational risk.
Explanation
Correct Answer
A:Cross-account access requires the provider to grant the intended objects and containing namespace appropriately. The stated acceptance evidence is: A consumer refreshes the imported share and can SELECT the newly granted secure view. Operational verification is stronger than an assumption based on saving a configuration.
Incorrect Answers
B:Zero-copy clones are logically independent even though unchanged historical partitions may be shared initially. That explains the role of “Source table row counts remain stable while the clone reflects only the test insert”, but the present requirement calls for “A consumer refreshes the imported share and can SELECT the newly granted secure view”: Cross-account access requires the provider to grant the intended objects and containing namespace appropriately.
C:Cloning and replication solve different recovery boundaries; a local test clone is not geographic resilience. That explains the role of “A disaster-recovery drill can access the replicated secondary account after loss of the primary region”, but the present requirement calls for “A consumer refreshes the imported share and can SELECT the newly granted secure view”: Cross-account access requires the provider to grant the intended objects and containing namespace appropriately.
D:Transient tables omit Fail-safe and differ from permanent tables in their recovery characteristics. That explains the role of “Object metadata reflects transient type and the retention policy matches the accepted operational risk”, but the present requirement calls for “A consumer refreshes the imported share and can SELECT the newly granted secure view”: Cross-account access requires the provider to grant the intended objects and containing namespace appropriately.
Question 953
A deployment team has configured this Snowflake capability: Inspect the query execution profile and underlying warehouse resources rather than treating this as a metadata outage. After promoting the configuration to the intended environment, auditors want observable evidence rather than a successful Save click. Which test outcome most convincingly verifies the capability is functioning as intended?
- CURRENT_DATABASE confirms the expected value after the statement sequence.
- A model version is registered and an authorized inference call references the pinned version.
- The execution profile shows expensive operators and the warehouse's run state and spill metrics.
- A load test demonstrates model inference latency, permissions and measured resource consumption.
Explanation
Correct Answer
C:Joins and scans consume warehouse processing resources; healthy session services do not guarantee fast execution. The stated acceptance evidence is: The execution profile shows expensive operators and the warehouse's run state and spill metrics. Operational verification is stronger than an assumption based on saving a configuration.
Incorrect Answers
A:An explicit session command changes the active namespace used for later unqualified SQL references. That explains the role of “CURRENT_DATABASE confirms the expected value after the statement sequence”, but the present requirement calls for “The execution profile shows expensive operators and the warehouse's run state and spill metrics”: Joins and scans consume warehouse processing resources; healthy session services do not guarantee fast execution.
B:A registry tracks model versions and deployment metadata so inference does not rely on ad-hoc file copies. That explains the role of “A model version is registered and an authorized inference call references the pinned version”, but the present requirement calls for “The execution profile shows expensive operators and the warehouse's run state and spill metrics”: Joins and scans consume warehouse processing resources; healthy session services do not guarantee fast execution.
D:Model serving architecture depends on the workload and supported services rather than automatically using standard SQL queries. That explains the role of “A load test demonstrates model inference latency, permissions and measured resource consumption”, but the present requirement calls for “The execution profile shows expensive operators and the warehouse's run state and spill metrics”: Joins and scans consume warehouse processing resources; healthy session services do not guarantee fast execution.
Question 954
A deployment team has configured this Snowflake capability: Grant USAGE and SELECT through reader roles while limiting MANAGE GRANTS and ownership authority. After promoting the configuration to the intended environment, auditors want observable evidence rather than a successful Save click. Which test outcome most convincingly verifies the capability is functioning as intended?
- Readers execute SELECT and attempts to grant privileged roles or alter objects fail.
- The reported configuration finding is documented, corrected and later shows resolved status.
- A support role sees the approved masked prefix and a privileged role sees the full value.
- Usage reporting attributes representative jobs to the intended cost centers.
Explanation
Correct Answer
A:Read-only business access should be separated from grant management and object modification privileges. The stated acceptance evidence is: Readers execute SELECT and attempts to grant privileged roles or alter objects fail. Operational verification is stronger than an assumption based on saving a configuration.
Incorrect Answers
B:Centralized posture views help detect configuration and security risks that object grants alone miss. That explains the role of “The reported configuration finding is documented, corrected and later shows resolved status”, but the present requirement calls for “Readers execute SELECT and attempts to grant privileged roles or alter objects fail”: Read-only business access should be separated from grant management and object modification privileges.
C:Column masking can transform visible values for less-privileged roles without changing stored source values. That explains the role of “A support role sees the approved masked prefix and a privileged role sees the full value”, but the present requirement calls for “Readers execute SELECT and attempts to grant privileged roles or alter objects fail”: Read-only business access should be separated from grant management and object modification privileges.
D:Clear compute ownership and query identification improve chargeback compared with account totals alone. That explains the role of “Usage reporting attributes representative jobs to the intended cost centers”, but the present requirement calls for “Readers execute SELECT and attempts to grant privileged roles or alter objects fail”: Read-only business access should be separated from grant management and object modification privileges.
Question 955
A deployment team has configured this Snowflake capability: Create or use a supported internal Snowflake stage with appropriate file format and grants. After promoting the configuration to the intended environment, auditors want observable evidence rather than a successful Save click. Which test outcome most convincingly verifies the capability is functioning as intended?
- An authorized query calls the endpoint and unapproved API hosts are rejected.
- The files are visible on the internal stage and the approved table load succeeds.
- Measured data age is checked against the target rather than assuming fixed refresh timestamps.
- A published Kafka test event is visible in Snowflake within the target freshness window.
Explanation
Correct Answer
B:Internal stages allow Snowflake-managed file staging without direct external cloud bucket access. The stated acceptance evidence is: The files are visible on the internal stage and the approved table load succeeds. Operational verification is stronger than an assumption based on saving a configuration.
Incorrect Answers
A:API integrations govern external-service invocation by Snowflake SQL functions under controlled permissions. That explains the role of “An authorized query calls the endpoint and unapproved API hosts are rejected”, but the present requirement calls for “The files are visible on the internal stage and the approved table load succeeds”: Internal stages allow Snowflake-managed file staging without direct external cloud bucket access.
C:Target lag guides Snowflake-managed refresh scheduling; it is not a guaranteed cron interval. That explains the role of “Measured data age is checked against the target rather than assuming fixed refresh timestamps”, but the present requirement calls for “The files are visible on the internal stage and the approved table load succeeds”: Internal stages allow Snowflake-managed file staging without direct external cloud bucket access.
D:The connector can route event rows to Snowflake's streaming ingestion instead of warehouse COPY batches. That explains the role of “A published Kafka test event is visible in Snowflake within the target freshness window”, but the present requirement calls for “The files are visible on the internal stage and the approved table load succeeds”: Internal stages allow Snowflake-managed file staging without direct external cloud bucket access.
Question 956
A deployment team has configured this Snowflake capability: Benchmark increasing warehouse size rather than adding multi-cluster capacity first. After promoting the configuration to the intended environment, auditors want observable evidence rather than a successful Save click. Which test outcome most convincingly verifies the capability is functioning as intended?
- Query execution time and billed credits are compared before and after resizing.
- The expected prior-month values appear and boundary rows have the defined NULL behavior.
- Representative JSON records yield the expected numeric and date values.
- A benchmark on representative JSON key lookups improves while returning identical rows.
Explanation
Correct Answer
A:Scale-up provides more resources for one running query, while extra clusters mainly serve concurrent queries. The stated acceptance evidence is: Query execution time and billed credits are compared before and after resizing. Operational verification is stronger than an assumption based on saving a configuration.
Incorrect Answers
B:LAG accesses preceding rows in a defined sequence without an expensive self-join for each offset. That explains the role of “The expected prior-month values appear and boundary rows have the defined NULL behavior”, but the present requirement calls for “Query execution time and billed credits are compared before and after resizing”: Scale-up provides more resources for one running query, while extra clusters mainly serve concurrent queries.
C:Typed extraction converts flexible input data into stable columns for reporting and validation. That explains the role of “Representative JSON records yield the expected numeric and date values”, but the present requirement calls for “Query execution time and billed credits are compared before and after resizing”: Scale-up provides more resources for one running query, while extra clusters mainly serve concurrent queries.
D:Eligible VARIANT field lookups can benefit from search optimization where supported by expression shape. That explains the role of “A benchmark on representative JSON key lookups improves while returning identical rows”, but the present requirement calls for “Query execution time and billed credits are compared before and after resizing”: Scale-up provides more resources for one running query, while extra clusters mainly serve concurrent queries.
Question 957
A deployment team has configured this Snowflake capability: Remove that consumer from the authorized share or revoke the relevant shared object grants. After promoting the configuration to the intended environment, auditors want observable evidence rather than a successful Save click. Which test outcome most convincingly verifies the capability is functioning as intended?
- Historical query output matches the expected pre-change or post-change rows around the recorded statement.
- A disaster-recovery drill can access the replicated secondary account after loss of the primary region.
- The table parameter displays the approved retention period and a historical test query succeeds.
- The former consumer can no longer query the share while the provider retains all source rows.
Explanation
Correct Answer
D:Provider-owned grants control consumer access independently of data persistence. The stated acceptance evidence is: The former consumer can no longer query the share while the provider retains all source rows. Operational verification is stronger than an assumption based on saving a configuration.
Incorrect Answers
A:Statement-based history offers precise temporal reference points for eligible Snowflake objects. That explains the role of “Historical query output matches the expected pre-change or post-change rows around the recorded statement”, but the present requirement calls for “The former consumer can no longer query the share while the provider retains all source rows”: Provider-owned grants control consumer access independently of data persistence.
B:Cloning and replication solve different recovery boundaries; a local test clone is not geographic resilience. That explains the role of “A disaster-recovery drill can access the replicated secondary account after loss of the primary region”, but the present requirement calls for “The former consumer can no longer query the share while the provider retains all source rows”: Provider-owned grants control consumer access independently of data persistence.
C:Extended retention may require Enterprise Edition or higher and incurs storage tradeoffs. That explains the role of “The table parameter displays the approved retention period and a historical test query succeeds”, but the present requirement calls for “The former consumer can no longer query the share while the provider retains all source rows”: Provider-owned grants control consumer access independently of data persistence.
Question 958
A deployment team has configured this Snowflake capability: Verify the active database and schema context used for unqualified object references. After promoting the configuration to the intended environment, auditors want observable evidence rather than a successful Save click. Which test outcome most convincingly verifies the capability is functioning as intended?
- Required controls are licensed and exercised successfully in the chosen edition.
- A notebook executes Python and SQL cells using the intended role and records reproducible outputs.
- Query history shows ETL statements running on the approved ETL warehouse.
- CURRENT_DATABASE and CURRENT_SCHEMA match the expected locations and the same query resolves the intended object.
Explanation
Correct Answer
D:Snowflake resolves unqualified names using session context and metadata rather than guessing tables from descriptions. The stated acceptance evidence is: CURRENT_DATABASE and CURRENT_SCHEMA match the expected locations and the same query resolves the intended object. Operational verification is stronger than an assumption based on saving a configuration.
Incorrect Answers
A:Feature availability varies by edition, and capabilities cannot be assumed merely because the SQL syntax exists. That explains the role of “Required controls are licensed and exercised successfully in the chosen edition”, but the present requirement calls for “CURRENT_DATABASE and CURRENT_SCHEMA match the expected locations and the same query resolves the intended object”: Snowflake resolves unqualified names using session context and metadata rather than guessing tables from descriptions.
B:Native notebooks support iterative code, SQL and visual analysis without requiring uncontrolled data exports. That explains the role of “A notebook executes Python and SQL cells using the intended role and records reproducible outputs”, but the present requirement calls for “CURRENT_DATABASE and CURRENT_SCHEMA match the expected locations and the same query resolves the intended object”: Snowflake resolves unqualified names using session context and metadata rather than guessing tables from descriptions.
C:Session warehouse context controls the compute resource charged for a query. That explains the role of “Query history shows ETL statements running on the approved ETL warehouse”, but the present requirement calls for “CURRENT_DATABASE and CURRENT_SCHEMA match the expected locations and the same query resolves the intended object”: Snowflake resolves unqualified names using session context and metadata rather than guessing tables from descriptions.
Question 959
A deployment team has configured this Snowflake capability: Audit PUBLIC role object grants and replace sensitive permissions with scoped custom roles. After promoting the configuration to the intended environment, auditors want observable evidence rather than a successful Save click. Which test outcome most convincingly verifies the capability is functioning as intended?
- SHOW GRANTS TO ROLE PUBLIC excludes sensitive objects and authorized custom roles retain access.
- The usage report shows monitored warehouse credits and independent serverless ingestion charges with actionable thresholds.
- A support role sees the approved masked prefix and a privileged role sees the full value.
- The report reproduces the seeded failed statement and its assigned warehouse and role.
Explanation
Correct Answer
A:Every user implicitly has PUBLIC, so sensitive grants there can create account-wide unintended access. The stated acceptance evidence is: SHOW GRANTS TO ROLE PUBLIC excludes sensitive objects and authorized custom roles retain access. Operational verification is stronger than an assumption based on saving a configuration.
Incorrect Answers
B:Warehouse resource monitors do not universally cap Snowflake-managed serverless feature charges, which require separate usage tracking. That explains the role of “The usage report shows monitored warehouse credits and independent serverless ingestion charges with actionable thresholds”, but the present requirement calls for “SHOW GRANTS TO ROLE PUBLIC excludes sensitive objects and authorized custom roles retain access”: Every user implicitly has PUBLIC, so sensitive grants there can create account-wide unintended access.
C:Column masking can transform visible values for less-privileged roles without changing stored source values. That explains the role of “A support role sees the approved masked prefix and a privileged role sees the full value”, but the present requirement calls for “SHOW GRANTS TO ROLE PUBLIC excludes sensitive objects and authorized custom roles retain access”: Every user implicitly has PUBLIC, so sensitive grants there can create account-wide unintended access.
D:Query metadata supports operational reporting of status, user, role and execution context. That explains the role of “The report reproduces the seeded failed statement and its assigned warehouse and role”, but the present requirement calls for “SHOW GRANTS TO ROLE PUBLIC excludes sensitive objects and authorized custom roles retain access”: Every user implicitly has PUBLIC, so sensitive grants there can create account-wide unintended access.
Question 960
A deployment team has configured this Snowflake capability: Define the correct PARQUET file format and compatible target column mapping. After promoting the configuration to the intended environment, auditors want observable evidence rather than a successful Save click. Which test outcome most convincingly verifies the capability is functioning as intended?
- An update to the downstream consumer causes the intermediate table to refresh in dependency order.
- The monitoring report includes a seeded failed refresh and the measured lag for each table.
- A sample load preserves amounts and timestamps with expected Snowflake column types.
- Task history shows scheduled executions and expected output table updates.
Explanation
Correct Answer
C:Parquet carries structured column data and should not be parsed as plain delimited text. The stated acceptance evidence is: A sample load preserves amounts and timestamps with expected Snowflake column types. Operational verification is stronger than an assumption based on saving a configuration.
Incorrect Answers
A:DOWNSTREAM delegates refresh scheduling to dependent tables instead of maintaining an independent timer. That explains the role of “An update to the downstream consumer causes the intermediate table to refresh in dependency order”, but the present requirement calls for “A sample load preserves amounts and timestamps with expected Snowflake column types”: Parquet carries structured column data and should not be parsed as plain delimited text.
B:Snowflake exposes scheduling, freshness and failure signals independent from ordinary query result values. That explains the role of “The monitoring report includes a seeded failed refresh and the measured lag for each table”, but the present requirement calls for “A sample load preserves amounts and timestamps with expected Snowflake column types”: Parquet carries structured column data and should not be parsed as plain delimited text.
D:Tasks execute statements on defined schedules, so manual worksheet sessions are unnecessary. That explains the role of “Task history shows scheduled executions and expected output table updates”, but the present requirement calls for “A sample load preserves amounts and timestamps with expected Snowflake column types”: Parquet carries structured column data and should not be parsed as plain delimited text.