The SnowPro Core certification is the foundation credential for Snowflake. This practice bank covers the same domains as the official study guide: platform architecture, account access and security, performance, data loading and unloading, transformations, and data protection / sharing.
Expect questions on micro-partitions, virtual warehouse sizing and auto-suspend, COPY INTO vs Snowpipe, RBAC (SYSADMIN vs ACCOUNTADMIN), result cache vs warehouse cache, Time Travel vs Fail-safe, and zero-copy cloning. Each item includes a rationale and a link to Snowflake documentation.
Pair this quiz with the Snowflake certification prep guide, the Snowflake SQL cheat sheet, and the Snowflake article category. Open the interactive quiz on this page (JavaScript required) to score by topic.
DataEngineer Hub is an independent study resource and is not affiliated with or endorsed by Snowflake Inc. Confirm current exam price, question count, and passing score on Snowflake's official certification page before you book.
The items below are from this bank so a reader (and an ads crawler) can study real stems without running JavaScript. The interactive quiz on this page has the full set, timed exam mode, and per-topic scoring.
Why: Snowflake uses a unique multi-cluster, shared-data architecture consisting of three layers: Database Storage (centralized, compressed, columnar), Query Processing (virtual warehouses = compute clusters), and Cloud Services (metadata, authentication, query optimization, infrastructure management). These layers scale independently.
Why: Multi-cluster warehouses automatically add/remove clusters based on concurrency. Setting `MAX_CLUSTER_COUNT > 1` with `SCALING_POLICY = STANDARD` (or ECONOMY) lets Snowflake spin up additional clusters when queries queue up, then scale down when demand drops. Increasing size helps with query speed but not concurrency. AUTO_SUSPEND controls idle shutdown, not scaling.
Why: Snowflake's COPY INTO natively supports CSV, JSON, Avro, ORC, Parquet, and XML. Excel (.xlsx) is not natively supported — you must convert it to CSV or Parquet first. All semi-structured formats (JSON, Avro, ORC, Parquet, XML) can be loaded into a VARIANT column.
Why: For Enterprise Edition and higher, Time Travel retention can be set from 0 to 90 days for permanent tables via `DATA_RETENTION_TIME_IN_DAYS`. Standard Edition is limited to 0 or 1 day. Transient and temporary tables are capped at 0 or 1 day regardless of edition. After Time Travel ends, Fail-safe begins (7 days, permanent tables only, not user-accessible).
Why: ACCOUNTADMIN is the most privileged role in a Snowflake account — it encompasses SYSADMIN + SECURITYADMIN and can view all billing/usage data. Best practice is to grant it only to a small number of users, enforce MFA, and do day-to-day work under SYSADMIN. ORGADMIN is at the organization level (across accounts) but has narrower scope within a single account.
Why: Snowflake uses zero-copy cloning: the clone only duplicates metadata pointers. No new storage is consumed until micro-partitions diverge due to DML on either the source or clone. This makes cloning instant and nearly free initially — ideal for dev/test environments.
Why: The Result Cache holds query results for 24 hours after execution. If the same query is re-run within that window, Snowflake serves it from cache (no warehouse needed). Each reuse extends the retention by another 24 hours, up to a maximum of 31 days. Underlying data changes invalidate the cache.
Why: Snowpipe is Snowflake-managed serverless continuous ingestion — triggered by cloud storage events (e.g., S3 SQS) or REST API, billed per-second of serverless compute. COPY INTO is a manual, batch, warehouse-driven load. Use Snowpipe for near-real-time micro-batches (<1 min latency); use COPY INTO for large scheduled bulk loads.
This preview is 8 of 100 questions. Use the quiz UI on this page to attempt the rest.
← Back to Home