ARA-C01 Question 292
Select 2Architecture changesA global retail company currently runs a single Snowflake account in AWS us-east-1 for all workloads. Due to a recent architecture review, the company must support European data residency requirements, reduce cross-region latency for EU analysts, and keep a subset of reference data synchronized between US and EU environments. The company also wants to minimize operational overhead and avoid rebuilding ingestion pipelines unless necessary. Which TWO architecture changes best meet these requirements?
- A
Create a separate Snowflake account in an EU region and use database replication and failover/failback features to synchronize the required reference databases from the US account.
- B
Create reader accounts for EU users in the existing us-east-1 account so queries run closer to the users without moving data.
- C
Deploy a separate Snowflake account in an EU region for EU-resident workloads and use Secure Data Sharing from the US account to the EU account for the shared reference data.
- D
Keep a single account in us-east-1 and enable Tri-Secret Secure so data is considered resident in the EU while remaining in the US region.
- E
Migrate all workloads into one organization account and use virtual warehouses in different sizes to satisfy regional residency and latency requirements.
Show answer and explanation
Correct answers: A, C
Explanation
The key architectural requirement is to introduce a multi-account, multi-region Snowflake design. To meet EU data residency and latency objectives, workloads serving EU users should run in a Snowflake account deployed in an EU region. Snowflake best practices commonly recommend multiple accounts for isolation, governance, and region-specific needs. For synchronizing selected data between regions, architects typically evaluate Snowflake-native options such as database replication and failover groups for maintaining copies of databases, or Secure Data Sharing when live shared access is appropriate and supported by the deployment pattern. Reader accounts and warehouse changes do not address region placement. Likewise, security features such as Tri-Secret Secure improve encryption and key control but do not alter the physical storage location of data. Relevant Snowflake documentation areas include account and region deployment planning, Secure Data Sharing, replication of databases and shares, and failover/failback for business continuity.
- A. Correct.
Correct. Creating a separate account in an EU region is the appropriate architectural change to satisfy data residency and reduce latency for EU-based users. Snowflake database replication can be used to synchronize selected databases, such as reference data, across supported regions/accounts. Failover/failback capabilities are part of Snowflake's cross-region and cross-cloud business continuity architecture and can also support managed synchronization patterns for replicated databases. This approach aligns with a multi-account, multi-region architecture and minimizes the need to redesign every ingestion path immediately because only the required shared datasets need replication.
- B. Incorrect.
Incorrect. Reader accounts do not move compute execution to a different region; they are tied to the provider account's region and are intended for sharing data with consumers who do not have their own Snowflake account. They do not solve EU data residency requirements or reduce cross-region data access latency in the way a regional deployment would. This option reflects the misconception that reader accounts provide geographic distribution of workloads.
- C. Correct.
Correct. A separate EU account is the right architectural foundation for residency and low-latency access. Secure Data Sharing is a valid way to share data between Snowflake accounts without copying the data, provided the sharing pattern and regional constraints are acceptable. For reference data that must be available in both environments, this can reduce operational overhead compared with rebuilding pipelines. In practice, architects choose between replication and sharing based on write requirements, regional boundaries, and whether physically duplicated copies are needed. This option is valid because it establishes the proper regional architecture and leverages a Snowflake-native mechanism to expose shared datasets with low management overhead.
- D. Incorrect.
Incorrect. Tri-Secret Secure enhances key management by adding a customer-managed key layer, but it does not change the physical region where data is stored. Data residency is determined by the Snowflake account's cloud platform region. Therefore, enabling Tri-Secret Secure in a US region does not make the data EU-resident. This is a common confusion between encryption control and geographic residency.
- E. Incorrect.
Incorrect. Warehouse sizing affects compute performance, not data residency or geographic proximity. A single account in us-east-1 still stores and processes data in that region. Even if the organization structure is adjusted, that does not by itself create a compliant EU-resident deployment. This option reflects the misconception that logical organization or compute scaling can replace a multi-region account architecture.