SnowPro Associate: Platform Question 15
Single answer● Snowflake layersA retail company is onboarding analysts to Snowflake and wants to explain why multiple business units can query the same sales data at the same time without slowing down nightly ELT jobs. The company stores all table data centrally and provisions separate virtual warehouses for ELT, finance reporting, and ad hoc analytics. Which Snowflake architecture component best explains this behavior?
- A
The database storage layer is separate from the compute layer, so multiple virtual warehouses can access the same centralized data independently
- B
Each virtual warehouse stores its own copy of table data in local storage, which prevents contention between workloads
- C
The cloud services layer executes all SQL queries directly, so warehouse size does not affect concurrency
- D
Snowflake schema design automatically isolates workloads by assigning each schema to a dedicated processing node
Show answer and explanation
Correct answer: A
Explanation
Snowflake's architecture consists of three main layers: database storage, compute, and cloud services. The key concept tested here is the separation of storage and compute. Persisted data is stored centrally in the storage layer, while one or more virtual warehouses in the compute layer can independently process queries against that same data. This design allows organizations to isolate workloads such as ETL, BI reporting, and exploratory analytics by using separate warehouses, improving concurrency and reducing resource contention. The cloud services layer supports orchestration functions such as metadata management, security, optimization, and transaction coordination, but it does not perform warehouse-based query execution itself. This aligns with Snowflake documentation describing its multi-cluster shared data architecture and the independent scaling of compute and storage.
- A. Correct.
Correct. Snowflake separates storage and compute. Data is stored once in centralized storage, while virtual warehouses provide independent compute resources. Because each warehouse has its own compute cluster, different workloads such as ELT and reporting can run concurrently against the same data with minimal resource contention.
- B. Incorrect.
Incorrect. Virtual warehouses do not maintain separate full copies of table data as part of normal Snowflake architecture. Snowflake uses centralized storage for persisted table data. Choosing this option reflects a common misconception based on traditional systems that tightly couple compute and storage.
- C. Incorrect.
Incorrect. The cloud services layer coordinates activities such as authentication, metadata management, query parsing, and optimization, but it does not replace virtual warehouses for query execution. Query processing still relies on compute resources in virtual warehouses for most data operations.
- D. Incorrect.
Incorrect. A schema is a logical container for database objects and does not provide workload isolation through dedicated processing nodes. Workload isolation in Snowflake is achieved primarily by assigning workloads to different virtual warehouses, not by schema design.