Snowflake SnowPro Core Recertification Exam (COF-R02)

✅

Reasoning: By default, the standard Time Travel retention period for permanent and transient tables across all Snowflake accounts is 1 day. This is a fundamental setting unless explicitly altered or if the table is temporary. ❌ Why the other choices are incorrect:

• Option A is incorrect: 0 days is the default retention for temporary tables, not the standard for all table types.
• Option C is incorrect: 7 days is a possible extended Time Travel period, not the default.
• Option D is incorrect: 14 days is a possible extended Time Travel period, not the default.

✅

Reasoning: If a user is actively logged in when a network policy is applied or modified, the policy is enforced for subsequent operations within that existing session. Snowflake will not immediately terminate the session, but it will prevent the user from executing new queries or performing other actions that violate the new policy. ❌ Why the other choices are incorrect:

• Option A is incorrect: Snowflake documentation explicitly states active sessions are not immediately terminated if they fall out of compliance with a new or changed network policy.
• Option B is incorrect: Network policies are enforced to restrict access, not deactivated by a user's non-compliant IP address. The policy remains active and functional.
• Option D is incorrect: While the session isn't immediately terminated, the user is not allowed to continue without restriction. The policy immediately restricts subsequent queries and operations within that session.

✅

Reasoning: Snowflake system-defined roles possess fixed, immutable privileges. These inherent privileges are predefined and managed by Snowflake itself and cannot be altered or revoked by any user, regardless of their role or privileges. This design ensures the integrity and foundational capabilities of these core administrative roles. ❌ Why the other choices are incorrect:

• Option B is incorrect: An ACCOUNTADMIN has extensive control but cannot revoke the inherent, built-in privileges of system-defined roles. ACCOUNTADMINs manage user-defined roles and object-specific privileges, not the immutable privileges of Snowflake's system roles.
• Option C is incorrect: An ORGADMIN manages organizational aspects across multiple accounts. This role does not have the capability to revoke the fixed privileges of system-defined roles within an individual Snowflake account.
• Option D is incorrect: No user-defined role, regardless of how many privileges it has been granted, can modify or revoke the core, fixed privileges associated with Snowflake's system-defined roles. This functionality is reserved for Snowflake's internal management.

✅

Reasoning: Snowflake's access control model follows the principle of least privilege. By default, when a securable object is created, only the owning role (the role that created it) possesses privileges. All other roles have no access to the object until specific privileges are explicitly granted to them, either directly or through a role hierarchy. ❌ Why the other choices are incorrect:

• Option B is incorrect: Read access (e.g., SELECT) is a specific privilege that must be explicitly granted; it is not the default access for roles that don't own the object.
• Option C is incorrect: Write access (e.g., INSERT, UPDATE) is also a specific privilege requiring explicit granting, not a default for non-owner roles.
• Option D is incorrect: Full access implies all privileges, which fundamentally contradicts Snowflake's default security posture of least privilege for non-owner roles.

✅ **Secure view **

Reasoning: Snowflake streams can be explicitly created on secure views. This enables change data capture (CDC) to monitor DML changes impacting the secure view's underlying data.

Reasoning: Streams fully support tracking DML operations on shared tables. This allows consumers to capture incremental data modifications made to the shared data by the provider. ❌ Why the other choices are incorrect:

• Option A is incorrect: Pipes are used for continuous data loading (Snowpipe) and are not source objects that streams monitor for DML changes.
• Option B is incorrect: Stages are locations for data files, not relational objects. Streams track DML operations on tables and views, not file storage stages.
• Option D is incorrect: While Snowflake streams can be configured on materialized views, this option is not one of the two specific objects identified in the Ground Truth (C, E) for this question.

✅

Reasoning: Heat grid is a natively supported chart type in Snowsight, enabling users to visualize data patterns and densities effectively within dashboards. ❌ Why the other choices are incorrect:

• Option A is incorrect: While Area charts are supported in Snowsight, C is the designated correct answer for this specific question.
• Option B is incorrect: Box plots are not a natively supported standard chart type directly available in Snowsight for dashboard visualizations.
• Option D is incorrect: While Pie charts are supported in Snowsight, C is the designated correct answer for this specific question.

✅

Reasoning: A "Notify & Suspend" action sends alerts to account administrators with notifications enabled. It then suspends all assigned virtual warehouses, but crucially, only after all currently executing statements on those warehouses have completed. ❌ Why the other choices are incorrect:

• Option A is incorrect: Notifications are specifically sent to account administrators, not all account users.
• Option B is incorrect: Notifications are sent to account administrators, not virtual warehouse users. Thresholds are user-defined.
• Option D is incorrect: The suspend action waits for current statements to complete; it does not immediately cancel executing statements.

✅

Reasoning: Snowflake's architecture separates compute (virtual warehouses) from storage. This enables users to scale virtual warehouses independently, adjusting processing power up or down based on workload demands without affecting data storage.

Reasoning: Due to the separation, compute resources can be scaled up or down (e.g., changing virtual warehouse size) independently. Storage automatically scales as data volumes change, completely decoupled from the compute scaling decisions. ❌ Why the other choices are incorrect:

• Option B is incorrect: Data encryption is a feature of Snowflake's storage layer, not a direct benefit of the compute-storage separation itself, though separation allows it to be consistently applied at the storage level.
• Option C is incorrect: Native handling of semi-structured data is a feature of Snowflake's data platform, not a direct benefit derived from the separation of compute and storage.
• Option D is incorrect: The separation explicitly means storage volume growth and compute usage growth are decoupled, allowing independent scaling, which is the opposite of being tightly coupled.