ADA-C01 Question 265
Single answerIdentify use casesA global retail company runs a business-critical Snowflake account in AWS us-east-1. The platform team has been asked to improve resilience for a small set of operational lookup tables and security metadata used by downstream applications in another Snowflake account in AWS us-west-2. The requirements are:
- Data in the secondary account must be queryable if the primary account becomes unavailable.
- Ongoing data movement should require minimal administrative effort.
- The solution should preserve database objects and grants as much as possible.
- Near-real-time cross-region transactional consistency is not required.
Which Snowflake feature is the best fit for this use case?
- A
Configure database replication and failover groups between the primary and secondary accounts
- B
Create a data share from the primary account to the secondary account and rely on Secure Data Sharing for disaster recovery
- C
Use external tables over cloud storage replicated between regions so the secondary account can query the same files
- D
Create a reader account in us-west-2 and direct downstream applications there during an outage
Show answer and explanation
Correct answer: A
Explanation
The best answer is database replication and failover groups because the scenario is fundamentally a business continuity/disaster recovery use case, not a data sharing or file access use case. In Snowflake, replication enables supported databases and certain account-level objects to be copied from a primary account to one or more secondary accounts, including across supported regions. Failover groups extend this for broader disaster recovery needs by grouping and replicating supported account objects and enabling failover. This is the appropriate pattern when an organization wants a secondary account to continue serving queries during a primary-region outage with limited administrative effort.
Why the other options are less suitable:
- Secure Data Sharing is best for distributing live data to other accounts, not for surviving unavailability of the source account/region.
- External tables are better suited to querying files in object storage, not preserving Snowflake-native databases, metadata, and grants for DR.
- Reader accounts are for consumer access, not cross-region account resilience.
This maps to Snowflake best practices for identifying use cases: choose replication/failover for disaster recovery and business continuity; choose Secure Data Sharing for governed data distribution; choose external tables for lake/file-based access patterns. Relevant Snowflake documentation includes the sections on Replication and Failover, Failover Groups, and Secure Data Sharing.
- A. Correct.
Correct. Database replication and failover groups are designed for cross-region and cross-account business continuity and disaster recovery use cases. They support replicating supported objects from a primary account to a secondary account and allow failover when needed. This aligns with the requirements for queryable data in another region, minimal ongoing administration, and preservation of objects and grants where supported. Because the scenario does not require active-active, near-real-time transactional consistency, replication/failover is the appropriate Snowflake-native resilience pattern.
- B. Incorrect.
Incorrect. Secure Data Sharing is intended for live data access without copying data, but it is not a disaster recovery mechanism for account or regional outages. If the provider account or region is unavailable, the share does not solve the resilience requirement in another region. A candidate might choose this because sharing is low-maintenance, but it does not provide the cross-region recovery semantics required here.
- C. Incorrect.
Incorrect. External tables can expose data stored in cloud object storage, and cross-region storage replication can make files available elsewhere, but this approach does not preserve Snowflake database objects and grants in the same way as replication/failover. It also shifts operational burden to storage replication design and does not address security metadata and broader database object continuity well. This is a plausible workaround for raw file access, but it is not the best fit for Snowflake account resilience.
- D. Incorrect.
Incorrect. Reader accounts allow data consumers without their own Snowflake account to access shared data, but they are not a disaster recovery strategy for an organization's separate Snowflake account in another region. They also do not address replication of operational lookup tables and security metadata for an existing downstream Snowflake account. Someone might pick this if they confuse distribution/access use cases with business continuity use cases.