You are tasked with deploying a new data application to Snowflake. This application requires several schemas for staging, transformation, and reporting. What is the recommended approach to create thes

Question

Options

• AManually create the schemas in each environment using the Snowflake web UI, documenting the
• BUse a Snowflake client (e.g., SnowSQL) and a script to execute 'CREATE SCHEMA' statements,
• CUtilize a third-party lac tool (e.g., Terraform) to define the schemas as resources, managing their
• DCopy the DDL (Data Definition Language) scripts generated by the Snowflake UI after manually
• EUtilize Snowflake's data replication feature to replicate schemas between different environments

• A
  5% (2)
• B
  3% (1)
• C
  57% (21)
• D
  24% (9)
• E
  11% (4)

Explanation

Using Terraform (or a similar IaC tool like Pulumi) to define Snowflake schemas as code resources is the gold standard because it enables version-controlled, repeatable, and environment-consistent deployments - you define schemas once and apply them identically to DEV, TEST, and PROD with environment-specific variables. Option A fails because manual UI work is error-prone, undocumented in code, and impossible to audit or reproduce reliably. Option B (raw SQL scripts via SnowSQL) is better than manual work but lacks state management - Terraform tracks what exists and only applies diffs, while scripts re-run blindly and risk errors. Option D (copy-pasting DDL from the UI) is just a semi-automated version of the manual approach - it still lacks environment parity guarantees and version control discipline. Option E (data replication) is designed to sync data, not to provision infrastructure, so it's the wrong tool entirely.

Memory tip: Think "IaC = Immutable, auditable, Consistent" - whenever a question mentions multiple environments needing the same infrastructure setup, the answer is always the tool that manages state and drift, which points to Terraform/IaC over scripts or manual methods.

No community discussion yet for this question.