ARA-C01 Question 142
Single answerSharing across cloud providersA data provider runs its primary Snowflake account in AWS us-east-1 and must distribute a curated dataset to two external customers: Customer A uses Snowflake in Azure East US 2, and Customer B uses Snowflake in AWS eu-west-1. The provider wants both customers to query the data with the least operational overhead while ensuring the data remains read-only to consumers. Which approach should the architect recommend?
- A
Create a direct share from the provider account in AWS us-east-1 to both consumer accounts, because secure sharing works transparently across all regions and cloud platforms.
- B
Use database replication or failover to copy the provider database into the customers' accounts, then grant the customers read access to the replicated databases.
- C
Create listings in a private data exchange or directly share from separate provider accounts/replicas in the target regions and cloud platforms, because standard direct shares require the provider data to exist in the same region as the consumer and cross-cloud/cross-region sharing uses replication to a local region first.
- D
Export the dataset to external stages in each cloud, have each customer load the files into its own Snowflake account, and use masking policies to preserve the read-only requirement.
Show answer and explanation
Correct answer: C
Explanation
The key architectural point is that Snowflake sharing is simplest when provider and consumer are in the same region, but sharing across regions or cloud platforms requires the provider to first make the data available in the consumer's target region/cloud, commonly through database replication. Once the dataset exists in a provider-controlled account in that target location, the provider can share it directly or publish it through listings/data exchange mechanisms. This approach minimizes operational overhead compared with exporting files, and it preserves Snowflake's native read-only consumption model for consumers. Relevant Snowflake documentation and best practices cover Secure Data Sharing, replication across regions/clouds, and cross-region/cross-cloud sharing patterns via listings and replication-backed provider accounts.
- A. Incorrect.
Incorrect. Standard Secure Data Sharing is not universally transparent across all regions and clouds from a single source region. Direct shares require the shared data to be available in the same region as the consumer account. To serve consumers in another region or cloud, the provider typically uses database replication so the data exists in a Snowflake account in the target region/cloud, then shares or lists it from there.
- B. Incorrect.
Incorrect. Replication and failover are account-level/provider-side capabilities used to make databases available in another Snowflake account or region under the provider's control, not to place replicated databases directly into customer-owned accounts for sharing. Customers should consume shared data through shares or listings, which preserve read-only access and avoid operational complexity in the consumer accounts.
- C. Correct.
Correct. For consumers in different regions or cloud platforms, the provider should first make the data available in the relevant target region/cloud using Snowflake replication, then expose it via sharing or listings from that local provider presence. This keeps the operational model aligned with Snowflake-native sharing, preserves read-only access for consumers, and supports cross-region and cross-cloud distribution patterns.
- D. Incorrect.
Incorrect. Exporting and reloading files increases operational overhead, breaks the simplicity of Snowflake-native sharing, and does not inherently provide the same governed read-only sharing model. Masking policies address data visibility, not the core distribution problem. This option also introduces file movement, load processes, and additional storage management that the scenario explicitly seeks to avoid.