dbt-Analytics-Engineering問題集、dbt Labs実際の試験問題

質問 1

You've defined development, staging, and production targets in your dbt project. You want to prevent accidental execution of destructive model materialization strategies (like drop or truncate) on the production target. How could you achieve this?

A. Write a custom macro that only executes such operations on non-production targets.

B. Rely purely on thorough code reviews to prevent execution of these commands in production.

C. Implement a CI/CD process that includes checks to prevent specific dbt commands from running in the production environment

D. Create a separate profiles.yml file for production that disables dangerous materialization strategies.

正解: A,C

解説: (PassTest メンバーにのみ表示されます)

質問 2

A team member proposes directly joining multiple raw source tables into a single large model for convenience.
Which of the following is a potential long-term consequence of this approach?

A. Maintenance complexity increases as business logic or source tables change.

B. Incremental materialization strategies would no longer be applicable.

C. Your DAG might inadvertently develop circular dependencies.

D. The dbt test command might run slower due to the complex join logic.

正解: A

解説: (PassTest メンバーにのみ表示されます)

質問 3

"Sometimes columns that should have data are showing null values!" exclaims a colleague. What initial debugging steps would you suggest beyond examining the model logic itself?

A. Verify that any where clauses or filters in the model are not inadvertently too restrictive.

B. Check any recent changes to source definitions or transformations on upstream data.

C. Review if there are snapshots or merge strategies that might impact how the table is updated.

D. All of the above.

正解: D

解説: (PassTest メンバーにのみ表示されます)

質問 4

You're onboarding a new team member and need to set them up with a personal sandbox environment within your existing project. What factors do you need to consider?

A. How to configure a profile on their machine and guide them to isolate changes to their sandbox.

B. Whether they need access to production data replicas and how to control this securely

C. All of the above.

D. How to provision data warehouse resources if needed, such as a separate schema or potentially even a dedicated database.

正解: C

解説: (PassTest メンバーにのみ表示されます)

質問 5

You need to grant readonly access to the results of your dbt project to stakeholders outside your team. Considering security, which approach would be the MOST suitable?

A. Embed dbt model results within a Bl dashboard and manage access through the Bl tool.

B. Create a designated reporting schema and set up views that map to the dbt model outputs

C. Have stakeholders directly query the dbt models, providing access to the production data warehouse.

D. Regularly run dbt seed to export CSV files, which are then shared with stakeholders.

正解: A,B

解説: (PassTest メンバーにのみ表示されます)

質問 6

The dataset you work with contains personally identifiable information (PII). Which types of additional tooling might be necessary for your project beyond a core dbt implementation?

A. Data masking or tokenization solutions.

B. Automated ETL pipelines

C. Enhanced data access governance and auditing tools.

D. Advanced statistical modeling tools.

正解: A,C

解説: (PassTest メンバーにのみ表示されます)

質問 7

A downstream process relies on a critical model to be fully refreshed daily. How can you enforce this behavior using dbt_project.yml configuration?

A. Update the model configuration to materialized: table and add a persist_docs: true property.

B. Set the model's materialized property to incremental _ refresh-

C. Configure a schedule property within the models: section.

D. Apply a post-hook that triggers a full refresh of the model.

正解: D

解説: (PassTest メンバーにのみ表示されます)

質問 8

Data Quality Issue

A. Add specific schema tests to the model, asserting that the column in question should not contain nulls.

B. Run dbt test on the model and see if existing tests fail, indicating a possible regression.

C. Run dbt compile followed by dbt run -select model_name to isolate the model and observe its output closely.

D. Inspect any filters or transformations within the model's SQL that might be unintentionally excluding data.

正解: A,B,C,D

解説: (PassTest メンバーにのみ表示されます)

質問 9

During development, you frequently used temporary print statements for debugging. Which of the following is the best practice before opening a pull request?

A. Consolidate the print statements into a separate debugging function that can be toggled on and off

B. Remove the print statements entirely from your commits-

C. Leave the print statements for future reference, in case similar debugging is needed-

D. Comment out the print statements but leave them in the code.

正解: B

解説: (PassTest メンバーにのみ表示されます)

質問 10

Logic Error

A. Adjust the model's SQL query and re-run it to observe the changes in output.

B. Create a separate branch specifically for addressing this issue.

C. Write new unit tests that specifically target the faulty calculation logic with various input scenarios.

D. Deploy the modified model to a development or staging environment.

正解: A

解説: (PassTest メンバーにのみ表示されます)

質問 11

You're experimenting with new transformations. You want to iterate on a model without affecting data in the production warehouse. Which dbt command should you use along with appropriate flags?

A. dbt compile to ensure changes are valid without executing models.

B. dbt seed with -select to load data into production.

C. dbt run with -target to specify a development environment.

D. dbt test with -defer to temporarily bypass upstream dependencies.

正解: C

解説: (PassTest メンバーにのみ表示されます)

質問 12

You're integrating a new dbt package and find its models conflict with naming conventions in your existing project. Which approach generally provides the most flexible solution?

A. Override the package's macros and redefine them within your project.

B. Contact the package maintainers to request changes aligning with your conventions.

C. Manually edit the SQL within the conflicting package models.

D. Utilize the dbt_project.yml file to re-alias the models provided by the package.

正解: D

解説: (PassTest メンバーにのみ表示されます)

dbt Labs dbt Analytics Engineering Certification - dbt-Analytics-Engineering 模擬練習