dbt-Analytics-Engineering Exam Dumps : dbt Analytics Engineering Certification Exam

In dbt, Jinja runs at compile time and operates only on Python objects that exist in the Jinja context (variables, lists, dictionaries, etc.). The tasks inside your with clause:

with tasks as (

select * from {{ ref('stg_tasks') }}

)

defines a SQL CTE named tasks, not a Jinja variable. Jinja cannot iterate over a SQL CTE; it can only loop over a Python iterable that has been created or passed into the template. Because there is no Jinja variable named tasks defined with something like:

{% set tasks = ['task_a', 'task_b', 'task_c'] %}

the for task in tasks loop has nothing to iterate over. As a result, the entire loop body is effectively skipped during compilation, and the compiled SQL only contains:

select

user_id,

with no generated case when expressions.

Option A is incorrect because loop.last is only needed for formatting (e.g., commas), not for the loop to render. Option B is wrong because Jinja control structures correctly use {% %}, while {{ }} is for output. Option D is irrelevant to compilation; missing columns would cause a runtime database error, not an empty compiled block.

Therefore, the problem is that no Jinja variable tasks was defined, making C the correct answer.

In dbt’s recommended layered modeling architecture, the staging layer is intended to provide a clean, one-to-one representation of raw source tables. Each stg_ model should depend only on sources, not on other staging models. This ensures staging remains a simple, transparent layer where data is renamed, recast, standardized, and lightly transformed before being passed to intermediate and mart layers.

In the DAG shown, at least one staging model (for example, stg_line_items or stg_tpch_line_items) appears downstream of another staging model, meaning a stg_ model is referencing another stg_ model. This violates dbt’s recommended modeling practice, because it creates unnecessary complexity in the staging layer and reduces modularity and transparency. Downstream layers such as intermediate (int_) and marts (fct_) should be used to combine, enrich, or join multiple staging outputs.

When staging models depend on each other, it becomes harder to trace lineage, reduces clarity about where transformations occur, and complicates the entire DAG. The proper pattern is:

Sources → Staging (stg_) → Intermediate (int_) → Marts (fct_)

Since the DAG shows staging models referencing other staging models, the rules have indeed been violated.

The correct answers are B and E.

The dbt retry command was introduced to help developers quickly re-run only what failed in the prior dbt invocation. According to dbt documentation, dbt retry examines metadata from the previous execution, including the run results and the manifest.json, to determine which nodes failed. This aligns with Option E, as dbt uses the manifest to identify the models, tests, and other nodes that errored or were not executed successfully.

Option B is also correct because dbt retry works only when the previous run failed due to execution errors (e.g., a warehouse failure or a logic error in SQL). If the failure was due to a syntax error that prevented dbt from compiling, there will be no valid manifest produced, so retry cannot execute. Thus, dbt retry can only re-run models when the previous failure occurred after successful compilation.

Options A and C are incorrect because retry does not run all nodes nor does it automatically pick up exactly where an execution stopped — instead, it runs failed and downstream unbuilt nodes as determined by the manifest. Option D is also incorrect; retry uses the execution metadata, not selectors, to determine what should run.