Snowflake

Snowflake Advanced Practice Questions: Architect, Data Engineer & Administrator

2026-03-21

更新: 2026-03-27

NicheeLab Editorial Team

The SnowPro Advanced exams (Architect / Data Engineer / Administrator) test advanced Snowflake design, optimization, and operations skills. This article walks through hand-picked practice questions from the Architect, Data Engineer, and Administrator exams, explaining the question patterns and the keys to picking the right answer.

Advanced questions are longer than Core questions and require design decisions that weigh multiple conditions at once. Read each explanation carefully and make sure you understand why that design is the best fit.

Comparing the Advanced Exams

Exam	Main Domains	Target Role	Key Knowledge Areas
Architect	Security architecture, Data Sharing, replication	Solutions Architects, infrastructure designers	Tri-Secret Secure, PrivateLink, failover, Organization
Data Engineer	Pipeline design, CDC, performance optimization	Data Engineers, ETL developers	Streams & Tasks, Snowpipe Streaming, MERGE INTO, Dynamic Tables
Administrator	Account administration, RBAC, cost management, monitoring	Platform administrators, DBAs	Resource Monitor, Network Policy, audit logs, ACCOUNT_USAGE

Architect Practice Question

SnowPro Advanced: Architect tests deep knowledge of platform design, security architecture, Data Sharing, and multi-region design. You need to know exactly which features each edition unlocks — particularly the ones specific to Business Critical.

Architect - Security Architecture

問題 1

A financial-services compliance requirement says Snowflake data must be protected with a customer-managed encryption key, and that revoking the key on the customer side must also stop Snowflake from accessing the data. Which combination of Snowflake feature and edition meets this requirement?

Enterprise Edition's standard AES-256 encryption combined with automatic key rotation satisfies the requirement
Use Tri-Secret Secure on Business Critical Edition, combining a customer-managed key with Snowflake's managed key
Import a custom encryption key on Standard Edition and use it
Only Virtual Private Snowflake (VPS) can meet this requirement

正解: B

Tri-Secret Secure combines a Snowflake-managed encryption key with a customer-managed key held in a cloud provider's KMS (AWS KMS, Azure Key Vault, or Google Cloud KMS). Because two independent keys are used, Snowflake cannot decrypt the data on its own. If the customer disables or deletes the key in their KMS, Snowflake's access to the data also stops. The feature is available on Business Critical Edition and above. AES-256 standard encryption (A) applies on every edition, but the key is Snowflake-managed and therefore doesn't satisfy the customer-managed-key requirement. Importing a custom key on Standard Edition (C) is not supported. Tri-Secret Secure is also available on VPS (D), but Business Critical is sufficient.

Data Engineer Practice Question

SnowPro Advanced: Data Engineer tests advanced knowledge of pipeline design, Snowpipe Streaming, streams and tasks, and performance tuning. CDC (Change Data Capture) implementation patterns are especially important.

Data Engineer - CDC Design

問題 2

In a large-scale ELT pipeline, you need to apply source-system change data (INSERT/UPDATE/DELETE) to a Snowflake target table at near real-time frequency. Which architecture is the most efficient?

Run a task on an hourly CRON expression that TRUNCATEs the source table and full-refreshes via COPY INTO
Ingest data into a staging table with Snowpipe Streaming, track the change delta with a stream, and run MERGE INTO on a scheduled task to update the target table
Export the data to an external stage and run COPY INTO manually once a day
Create Dynamic Tables on top of the staging table and rely on automatic refresh to update the final table

正解: B

For a CDC (Change Data Capture) pipeline that reflects source INSERT/UPDATE/DELETE in near real time, the combination of Snowpipe Streaming + streams & tasks + MERGE INTO is the most efficient. Snowpipe Streaming lands data in a staging table with seconds-level latency, a standard stream tracks INSERT/UPDATE/DELETE on the staging table, and a task periodically runs MERGE INTO to apply the delta to the target. Full refresh (A) is inefficient at scale, and manual loading (C) doesn't meet the near real-time requirement. Dynamic Tables (D) are great for INSERT-based transformations but aren't well suited to MERGE-style logic that propagates source UPDATEs/DELETEs to the target.

Administrator Practice Question

SnowPro Advanced: Administrator tests knowledge of account administration, RBAC design, cost management, monitoring, and auditing. Mastery of Resource Monitor and the ACCOUNT_USAGE views is especially important.

Administrator - Cost Management

問題 3

A Snowflake account's monthly credit consumption is significantly over budget. Investigation shows the cause is serverless compute from Snowpipe, Automatic Clustering, and Materialized View auto-maintenance. Which method should you use to analyze these costs in detail?

Configure a Resource Monitor to track serverless compute credit consumption
Analyze cost by service using METERING_HISTORY, AUTOMATIC_CLUSTERING_HISTORY, and MATERIALIZED_VIEW_REFRESH_HISTORY in the ACCOUNT_USAGE schema
Use SHOW WAREHOUSES to check credit consumption for every warehouse
Use the Snowsight Usage dashboard to view per-warehouse cost graphs

正解: B

Resource Monitor only tracks Virtual Warehouse credit consumption; it does not cover serverless compute such as Snowpipe, Automatic Clustering, or Materialized View auto-maintenance (so A is wrong). To analyze serverless compute cost in detail, use the ACCOUNT_USAGE views: METERING_HISTORY (Snowpipe), AUTOMATIC_CLUSTERING_HISTORY (clustering), MATERIALIZED_VIEW_REFRESH_HISTORY (Materialized View), and SEARCH_OPTIMIZATION_HISTORY (Search Optimization Service). These let you pinpoint credit consumption by service, table, and day. SHOW WAREHOUSES (C) and the Usage dashboard (D) are useful for inspecting warehouse consumption but aren't sufficient for detailed serverless analysis.

How to Beat the Advanced Exams

1. Get comfortable with scenario-based questions

Unlike Core, most Advanced questions are scenario-based: "What is the best design for this requirement?" or "What is causing this performance issue and how do you fix it?" Question stems are long, so you need the reading speed to quickly pick out the key constraints — edition, compliance requirements, data volume, update frequency, and latency requirements.

2. Know exactly which features each edition unlocks

Feature	Required Edition
Tri-Secret Secure	Business Critical or higher
Private Link (AWS/Azure/GCP)	Business Critical or higher
Database failover / failback	Business Critical or higher
Dynamic Data Masking	Enterprise or higher
Row Access Policy	Enterprise or higher
Automatic key rotation	Enterprise or higher
Search Optimization Service	Enterprise or higher
Materialized View	Enterprise or higher

3. Lean on real-world design experience

Many Advanced questions are hard to answer correctly without real-world experience. If your hands-on time is limited, focus your study on the "Best Practices" and "Considerations" sections of the official documentation. The most important thing to internalize is when each feature should — and should not — be used.

4. Build a clear cost-optimization mental model

Advanced exams frequently ask you to choose the most cost-efficient design. Be crystal-clear on the difference between warehouse consumption (user-managed) and serverless consumption (Snowflake-managed), and on the distinction between storage cost (capacity-based) and compute cost (credit-based).

Want to solve more questions?

Practice every Advanced exam domain comprehensively with the NicheeLab question bank

Try free questions →

Frequently Asked Questions

What are the prerequisites for the SnowPro Advanced exams?

Every SnowPro Advanced exam (Architect / Data Engineer / Administrator) requires a current SnowPro Core (COF-C03) certification as a prerequisite. Your Core certification must still be valid (within 2 years of earning it). Advanced exams have 65 questions in 115 minutes, a passing score of 750/1000, and a fee of $375. The question count is lower than Core, but each question is significantly harder and tests design decisions grounded in real-world experience.

Which Advanced exam should I take first: Architect, Data Engineer, or Administrator?

Choose based on your day job. Architect is right for people working on infrastructure design, security architecture, Data Sharing, and replication design. Data Engineer fits those building ETL/ELT pipelines, Snowpipe, streams and tasks, and performance tuning. Administrator is best if you handle account administration, RBAC design, cost management, and monitoring. All three assume at least 1 year of hands-on Snowflake experience.

How long should I plan to study for an Advanced exam?

With 1+ years of Snowflake experience, plan on 2-3 months (roughly 100-150 hours total). Build on your Core knowledge and go deep on each Advanced exam's specialty area. Architect requires Tri-Secret Secure, PrivateLink, and failover design. Data Engineer focuses on streams and tasks, Snowpipe Streaming, and CDC design. Administrator covers Resource Monitor, RBAC hierarchy design, and audit logs. If your hands-on experience is limited, budget 3-4 months instead.

Check what you learned with practice questions

Practice with certification-focused question sets

無料で問題を解いてみる

Author

NicheeLab Editorial Team

NicheeLab editorial team focused on data engineering and cloud certification learning. Content is structured around practical study needs and official exam domains.

Snowflake

Snowflake Certifications: All 11 Exams Explained (2026)

Every SnowPro certification — Associate, Core, Specialty, Ad...

Snowflake

Snowflake Exam Difficulty Ranking: All 11 Certs Compared (2026)

All 11 SnowPro exams ranked by difficulty with study-time es...

Snowflake

Snowflake Study Guide: Fastest Pass Route by Exam (2026)

How to pass SnowPro certifications efficiently — official ma...

Snowflake

SnowPro Core (COF-C03): Complete Exam Guide (2026)

Pass the SnowPro Core exam — six domains, scope, sample ques...