SnowPro Advanced: Security Engineer Question 285
Select 2Changes in serverless compute consumptionA Security Engineer notices that the ACCOUNTADMIN role has raised concerns about a sudden increase in daily costs attributed to serverless compute after several new governance controls were deployed. The team recently enabled dynamic data masking on sensitive columns, added row access policies to high-volume tables, and scheduled additional automated maintenance tasks. They want to determine which recent changes are most likely to increase serverless compute consumption in Snowflake so they can investigate the right areas first. Which TWO changes are the most likely contributors?
- A
Increasing the frequency and number of serverless TASK executions used for governance and monitoring workflows
- B
Adding more row access and masking policies that must be evaluated during query processing on frequently accessed objects
- C
Resizing a user-managed virtual warehouse from MEDIUM to LARGE for ETL jobs
- D
Granting the APPLY MASKING POLICY privilege to additional security administrators
- E
Creating additional network policies to restrict login locations
Show answer and explanation
Correct answers: A, B
Explanation
In Snowflake, serverless compute consumption can rise when organizations expand features that rely on Snowflake-managed compute rather than customer-managed warehouses. Two practical areas to investigate first are: (1) serverless TASK activity, especially when tasks are scheduled more frequently for governance or monitoring, and (2) expanded use of row access policies and masking policies on heavily queried data, because policy enforcement can add to serverless services usage. By contrast, resizing a virtual warehouse affects warehouse credits, not serverless credits. Administrative privilege grants and creation of network policies are security changes, but they are not common drivers of recurring serverless compute growth. For exam purposes, candidates should distinguish between customer-managed warehouse consumption and Snowflake-managed serverless/services consumption, and recognize that governance features can influence the latter when applied broadly at scale.
- A. Correct.
Correct. Serverless TASKs use Snowflake-managed compute rather than customer-managed virtual warehouses. Increasing how often they run, or adding more of them, can directly raise serverless compute consumption. This is a common source of unexpected cost growth when teams automate governance, monitoring, or remediation workflows.
- B. Correct.
Correct. Row access policies and masking policies can contribute to serverless compute consumption because policy evaluation is part of Snowflake's serverless services layer. If these policies are attached to heavily queried objects, especially in high-volume environments, the associated serverless usage can increase. This is particularly relevant after broad rollout of governance controls across frequently accessed tables and columns.
- C. Incorrect.
Incorrect. Resizing a user-managed virtual warehouse affects warehouse compute consumption, not serverless compute consumption. While total account cost may rise, this change would appear under warehouse billing rather than serverless services usage.
- D. Incorrect.
Incorrect. Granting privileges changes authorization metadata but does not itself consume meaningful ongoing serverless compute. The misconception is confusing administrative configuration changes with recurring runtime evaluation or execution costs.
- E. Incorrect.
Incorrect. Network policies control authentication access restrictions, but simply creating more of them does not typically drive serverless compute consumption in the way recurring serverless tasks or policy evaluations can. They are important security controls, but not a likely cause of sustained increases in serverless usage.