100% Latest & Updated Snowflake SnowPro Core COF-C03 Practice Test Questions, Exam Dumps & Verified Answers!
30 Days Free Updates, Instant Download!
Question 10
A platform engineer proposes installing Snowflake server binaries on a private on-premises server Which technical action is most appropriate under this requirement?
- Copy warehouse binaries into an on-premises container
- Use the managed Snowflake service on a supported cloud and region
- Create an unmanaged database service on an ESXi host
- Install a local Snowflake database engine from a package manager
Explanation
Correct Answer
B:Use the managed Snowflake service on a supported cloud and region. This addresses the stated condition without bypassing the required boundary.
Incorrect Answers
A:The problem: A platform engineer proposes installing Snowflake server binaries on a private on-premises server. The appropriate decision: Use the managed Snowflake service on a supported cloud and region.
C:The problem: A platform engineer proposes installing Snowflake server binaries on a private on-premises server. The appropriate decision: Use the managed Snowflake service on a supported cloud and region.
D:The problem: A platform engineer proposes installing Snowflake server binaries on a private on-premises server. The appropriate decision: Use the managed Snowflake service on a supported cloud and region.
Question 11
An architecture plan requires a dedicated isolated ETL execution cluster and independent analyst compute Which proposal best pairs a technically correct change with the outcome that would validate it?
- Assign ETL and interactive work to separate warehouses with workload-specific sizing; verify Long transformations do not queue unrelated interactive queries
- Rename the ETL role without changing compute assignment; no direct validation of Long transformations do not queue unrelated interactive queries
- Create a second database but keep one warehouse; no direct validation of Long transformations do not queue unrelated interactive queries
- Add a clustering key to an unrelated small table; no direct validation of Long transformations do not queue unrelated interactive queries
Explanation
Correct Answer
A:Assign ETL and interactive work to separate warehouses with workload-specific sizing. Verify: Long transformations do not queue unrelated interactive queries.
Incorrect Answers
B:Required correction: Assign ETL and interactive work to separate warehouses with workload-specific sizing. Evidence: Long transformations do not queue unrelated interactive queries.
C:Required correction: Assign ETL and interactive work to separate warehouses with workload-specific sizing. Evidence: Long transformations do not queue unrelated interactive queries.
D:Required correction: Assign ETL and interactive work to separate warehouses with workload-specific sizing. Evidence: Long transformations do not queue unrelated interactive queries.
Question 12
Storage cost estimates assume every warehouse duplicates all table data Which technical action is most appropriate under this requirement?
- Copy every base table into a separate schema per warehouse
- Set MAX_CLUSTER_COUNT equal to stored terabytes
- Distinguish centrally maintained compressed table storage from independent warehouse compute
- Keep one warehouse for all users regardless of concurrency
Explanation
Correct Answer
C:Distinguish centrally maintained compressed table storage from independent warehouse compute. This addresses the stated condition without bypassing the required boundary.
Incorrect Answers
A:The problem: Storage cost estimates assume every warehouse duplicates all table data. The appropriate decision: Distinguish centrally maintained compressed table storage from independent warehouse compute.
B:The problem: Storage cost estimates assume every warehouse duplicates all table data. The appropriate decision: Distinguish centrally maintained compressed table storage from independent warehouse compute.
D:The problem: Storage cost estimates assume every warehouse duplicates all table data. The appropriate decision: Distinguish centrally maintained compressed table storage from independent warehouse compute.
Question 13
A JSON array contains multiple line items and a report needs one output row per element Which technical action is most appropriate under this requirement?
- Call an aggregate COUNT without expanding the array
- Use LATERAL FLATTEN on the array expression
- Turn the JSON into a network policy
- Replace FLATTEN with warehouse auto-resume
Explanation
Correct Answer
B:Use LATERAL FLATTEN on the array expression. This addresses the stated condition without bypassing the required boundary.
Incorrect Answers
A:The problem: A JSON array contains multiple line items and a report needs one output row per element. The appropriate decision: Use LATERAL FLATTEN on the array expression.
C:The problem: A JSON array contains multiple line items and a report needs one output row per element. The appropriate decision: Use LATERAL FLATTEN on the array expression.
D:The problem: A JSON array contains multiple line items and a report needs one output row per element. The appropriate decision: Use LATERAL FLATTEN on the array expression.
Question 14
A marketplace data product includes contractual usage constraints and must restrict who can see sensitive fields Select the THREE statements identifying the correct action, direct evidence, and unresolved release-blocking risk. (Choose THREE.)
- Grant PUBLIC on every sensitive provider source
- Authorized consumers see the intended schema and restricted fields stay protected
- Disable masking to simplify the listing demo
- Block release while: Marketing listing visibility is mistaken for unrestricted base table visibility
- Share a database owner password as the access method
- Publish the product through a governed listing with applicable provider security policies
Explanation
Correct Answers
B:Acceptance evidence: Authorized consumers see the intended schema and restricted fields stay protected. Completing configuration alone is insufficient.
D:Unresolved failure: Marketing listing visibility is mistaken for unrestricted base table visibility. Correct and test it before acceptance.
F:Publish the product through a governed listing with applicable provider security policies. This addresses the stated condition without bypassing the required boundary.
Incorrect Answers
A:The problem: A marketplace data product includes contractual usage constraints and must restrict who can see sensitive fields. The appropriate decision: Publish the product through a governed listing with applicable provider security policies.
C:The problem: A marketplace data product includes contractual usage constraints and must restrict who can see sensitive fields. The appropriate decision: Publish the product through a governed listing with applicable provider security policies.
E:The problem: A marketplace data product includes contractual usage constraints and must restrict who can see sensitive fields. The appropriate decision: Publish the product through a governed listing with applicable provider security policies.
Question 15
Sensitive columns must be classified and labeled consistently for governance reporting Which technical action is most appropriate under this requirement?
- Disable ACCOUNT_USAGE views to avoid classification work
- Set a table comment only and assume it enforces the tag policy
- Use supported classification and tags with defined taxonomy and ownership
- Rename every warehouse to a sensitivity category
Explanation
Correct Answer
C:Use supported classification and tags with defined taxonomy and ownership. This addresses the stated condition without bypassing the required boundary.
Incorrect Answers
A:The problem: Sensitive columns must be classified and labeled consistently for governance reporting. The appropriate decision: Use supported classification and tags with defined taxonomy and ownership.
B:The problem: Sensitive columns must be classified and labeled consistently for governance reporting. The appropriate decision: Use supported classification and tags with defined taxonomy and ownership.
D:The problem: Sensitive columns must be classified and labeled consistently for governance reporting. The appropriate decision: Use supported classification and tags with defined taxonomy and ownership.
Question 16
A team develops a distributable application with logic that customers install inside their own Snowflake accounts Select the TWO choices that together provide a supported correction and directly relevant acceptance evidence. (Choose TWO.)
- Require buyers to edit provider production procedures directly
- Assess the Snowflake Native App Framework and application package model
- Ship privileged account credentials in the package
- A consumer can install a controlled package with the intended permissions and functionality
- Use a plain database share for every executable workflow
Explanation
Correct Answers
B:Assess the Snowflake Native App Framework and application package model. This addresses the stated condition without bypassing the required boundary.
D:Acceptance evidence: A consumer can install a controlled package with the intended permissions and functionality. Completing configuration alone is insufficient.
Incorrect Answers
A:The problem: A team develops a distributable application with logic that customers install inside their own Snowflake accounts. The appropriate decision: Assess the Snowflake Native App Framework and application package model.
C:The problem: A team develops a distributable application with logic that customers install inside their own Snowflake accounts. The appropriate decision: Assess the Snowflake Native App Framework and application package model.
E:The problem: A team develops a distributable application with logic that customers install inside their own Snowflake accounts. The appropriate decision: Assess the Snowflake Native App Framework and application package model.
Question 17
An analyst needs to run repeatable SQL-and-Python exploration with shared project artifacts Which technical action is most appropriate under this requirement?
- Treat a notebook cell as a database role grant
- Replace reproducibility with copied screenshots
- Grant PUBLIC access to every source table
- Use Snowflake Notebooks with authorized roles and supported execution settings
Explanation
Correct Answer
D:Use Snowflake Notebooks with authorized roles and supported execution settings. This addresses the stated condition without bypassing the required boundary.
Incorrect Answers
A:The problem: An analyst needs to run repeatable SQL-and-Python exploration with shared project artifacts. The appropriate decision: Use Snowflake Notebooks with authorized roles and supported execution settings.
B:The problem: An analyst needs to run repeatable SQL-and-Python exploration with shared project artifacts. The appropriate decision: Use Snowflake Notebooks with authorized roles and supported execution settings.
C:The problem: An analyst needs to run repeatable SQL-and-Python exploration with shared project artifacts. The appropriate decision: Use Snowflake Notebooks with authorized roles and supported execution settings.
Question 18
A large multi-way JOIN produces billions of intermediate rows from a few million inputs Which unresolved finding would be a valid reason to block acceptance after the change?
- The team confirms: Intermediate row counts fall and the returned result still meets the data contract
- Change and rollback evidence is retained with the release
- An accidental many-to-many join explodes row counts and cost
- The relevant users retain exactly their approved access
Explanation
Correct Answer
C:Unresolved failure: An accidental many-to-many join explodes row counts and cost. Correct and test it before acceptance.
Incorrect Answers
A:This is a legitimate target or safeguard, not the unresolved failure described in the question: An accidental many-to-many join explodes row counts and cost.
B:“Change and rollback evidence is retained with the release” describes a target or safeguard, not the release-blocking failure. The defect: An accidental many-to-many join explodes row counts and cost.
D:“The relevant users retain exactly their approved access” describes a target or safeguard, not the release-blocking failure. The defect: An accidental many-to-many join explodes row counts and cost.
Question 19
An ETL role can select records but cannot run queries on its designated warehouse Select the TWO choices that together provide a supported correction and directly relevant acceptance evidence. (Choose TWO.)
- Transfer OWNERSHIP of every table to the ETL role
- Remove the database-level role
- Add a result-cache setting instead of warehouse USAGE
- The role runs its authorized query on the designated compute resource
- Grant the appropriate USAGE privilege on the intended warehouse to its active role
Explanation
Correct Answers
D:Acceptance evidence: The role runs its authorized query on the designated compute resource. Completing configuration alone is insufficient.
E:Grant the appropriate USAGE privilege on the intended warehouse to its active role. This addresses the stated condition without bypassing the required boundary.
Incorrect Answers
A:The problem: An ETL role can select records but cannot run queries on its designated warehouse. The appropriate decision: Grant the appropriate USAGE privilege on the intended warehouse to its active role.
B:The problem: An ETL role can select records but cannot run queries on its designated warehouse. The appropriate decision: Grant the appropriate USAGE privilege on the intended warehouse to its active role.
C:The problem: An ETL role can select records but cannot run queries on its designated warehouse. The appropriate decision: Grant the appropriate USAGE privilege on the intended warehouse to its active role.
Question 20
A reporting job has unpredictable execution length and operators confuse query timeout with warehouse idle timeout Select the TWO choices that together provide a supported correction and directly relevant acceptance evidence. (Choose TWO.)
- Put timeout controls in an unrelated stage definition
- Long-running statements and idle periods follow the distinct defined policies
- Set only STATEMENT_TIMEOUT and assume it suspends idle compute
- Configure statement timeout and auto-suspend independently for their separate purposes
- Use AUTO_SUSPEND as an application request timeout
Explanation
Correct Answers
B:Acceptance evidence: Long-running statements and idle periods follow the distinct defined policies. Completing configuration alone is insufficient.
D:Configure statement timeout and auto-suspend independently for their separate purposes. This addresses the stated condition without bypassing the required boundary.
Incorrect Answers
A:The problem: A reporting job has unpredictable execution length and operators confuse query timeout with warehouse idle timeout. The appropriate decision: Configure statement timeout and auto-suspend independently for their separate purposes.
C:The problem: A reporting job has unpredictable execution length and operators confuse query timeout with warehouse idle timeout. The appropriate decision: Configure statement timeout and auto-suspend independently for their separate purposes.
E:The problem: A reporting job has unpredictable execution length and operators confuse query timeout with warehouse idle timeout. The appropriate decision: Configure statement timeout and auto-suspend independently for their separate purposes.
SnowPro Core COF-C03 Practice Tests