Databricks Data Engineer Associate Working with Lakeflow Jobs: 129 practice questions

12 of the 129 Working with Lakeflow Jobs questions in the Certsqill Databricks Data Engineer Associate bank, shown in full below. Each one carries an explanation for every option, not just the correct one — the wrong answers are where the marks go.

Preparing for Databricks Data Engineer Associate? Take the free 5-min readiness check →

1. Every 15 minutes in UTC: What cadence does this schedule express?

1. Every 15 minutes in UTC. ✓

   The Quartz expression fires at minute zero and every fifteen-minute interval, with UTC explicitly selected.
2. Once every hour in UTC.

   An hourly schedule would not include the repeated fifteen-minute interval expression shown here.
3. Whenever a telemetry file arrives.

   File-arrival triggers are data-driven; this configuration is an explicit time-based schedule.
4. Whenever the telemetry source table changes.

   Table-update triggers respond to source updates, whereas the supplied configuration defines a recurring clock schedule.

The configured Quartz schedule runs at fifteen-minute intervals using UTC.

2. Deduplicate the target by ` `: Which change addresses both existing duplicates and future retries?

1. Disable retries and leave the existing target unchanged.

   This avoids one automatic repetition but neither repairs existing duplicates nor makes manual repairs safe.
2. Increase retries and retain the append write.

   Retries repeat the non-idempotent append and can create more duplicates.
3. Deduplicate the target by `(product, metric_date)`, deduplicate each batch by the same key, then MERGE the batch on that key. ✓

   Cleaning the target establishes one existing row per key, while source deduplication and keyed MERGE prevent duplicate effects from retries and repairs.
4. Replace append with a keyed MERGE from the source.

   A MERGE can make future writes idempotent only when the source is unique by key; by itself it does not remove duplicates already in the target.

Clean existing duplicates, deduplicate each batch, and use a keyed MERGE.

3. A SQL task containing the query: Which Lakeflow Jobs task type is most direct?

1. A file-arrival trigger.

   A trigger starts a run but does not execute the required SQL statement.
2. A Python script task that submits the SQL.

   This adds an unnecessary Python layer when the workload is already a standalone SQL statement.
3. A notebook task containing the SQL.

   A notebook can run SQL, but the scenario specifically needs direct SQL execution without a notebook wrapper.
4. A SQL task containing the query. ✓

   A configured SQL task directly executes the statement against the registered table without Python or notebook code.

Use a SQL task because the required workload is a standalone SQL statement.

4. A completion marker written after upload finishes: What prerequisite should the ingestion design add?

1. A shorter polling interval for the file trigger.

   More frequent polling detects arrivals sooner but cannot establish upload completeness.
2. A table-update trigger on an unrelated summary table.

   An unrelated table update does not prove that the multipart source file is complete.
3. A completion marker written after upload finishes. ✓

   A producer-controlled completion signal distinguishes a complete upload from a file still being written.
4. A larger job cluster to process the first piece faster.

   Additional compute may speed processing but cannot establish whether the supplier has finished uploading.

Use a completion marker so file arrival represents a finished upload.

5. A file-arrival trigger for the source location: Which trigger is the best fit?

1. A file-arrival trigger for the source location. ✓

   A file-arrival trigger responds to new files at the configured supported location, matching unpredictable delivery timing.
2. A table-update trigger on the downstream report.

   A downstream update occurs after processing and cannot reliably initiate processing of newly arriving source files.
3. An hourly trigger with a shorter timeout.

   A time trigger follows the clock and can add latency or unnecessary runs when delivery is unpredictable.
4. A daily trigger near the average delivery time.

   Average delivery timing is not guaranteed and may miss files delivered outside the selected time.

Use a file-arrival trigger when unpredictable source files should initiate processing.

6. Run remediation when at least one dependency fails: Which dependency behavior is appropriate?

1. Run remediation after all dependencies complete, regardless of outcome.

   A completion condition also runs after successful validation, violating the requirement that remediation run only on failure.
2. Remove the dependency and schedule remediation independently.

   Without the dependency, remediation can run before validation establishes a failure.
3. Run remediation when at least one dependency fails. ✓

   This preserves the dependency and selects the failure path when the validation task fails.
4. Run remediation only after all dependencies succeed.

   An all-success condition skips remediation when validation fails.

Configure remediation to run when at least one dependency fails.

7. The SQL task runs only after the pipeline task reaches: What behavior does this dependency specify?

1. The SQL task runs whenever the pipeline definition is edited.

   Editing pipeline source does not establish a runtime task dependency or automatically execute the downstream SQL task.
2. The SQL task runs only after the pipeline task reaches its required successful completion. ✓

   The dependency orders the tasks, and default successful dependency behavior allows quality checking after ingestion succeeds.
3. The pipeline task automatically infers that the SQL query is its downstream transformation.

   The explicit dependency is configured in the job; tasks do not automatically infer business dependencies from query meaning.
4. The SQL task runs concurrently so it can validate partial pipeline output.

   A depends_on edge orders execution rather than allowing concurrent reading of partial pipeline output.

The dependency orders the quality check after successful supplier-pipeline completion.

8. A weekday time trigger scheduled for 06: Which trigger best satisfies the requirement?

1. A table-update trigger on whichever telemetry table changes first.

   The first table update does not establish the required fixed start time or complete source freshness.
2. A file-arrival trigger on the latest telemetry export.

   File arrival timing depends on delivery and cannot guarantee a predictable weekday 06:00 UTC start.
3. A table-update trigger requiring all telemetry tables to change first.

   Waiting for table updates can delay or prevent the run when sources update at different times.
4. A weekday time trigger scheduled for 06:00 UTC. ✓

   A time trigger starts on the required predictable schedule, after which freshness validation can determine data suitability.

Use a weekday time trigger for a fixed reporting deadline, then validate source freshness separately.

9. Use an all-table update trigger for the three source: Which implementation satisfies these requirements?

1. Use a file-arrival trigger on the log-file directory.

   File arrival at one location does not establish that all three Unity Catalog tables have updated.
2. Use an all-table update trigger for the three source tables. ✓

   An all-table update trigger waits for updates from each configured source before starting the job.
3. Use a table-update trigger for any one of the three source tables and let transformation start after that event.

   An any-table condition can start while the other two tables remain stale, violating the all-source requirement.
4. Use a scheduled trigger for the transformation.

   A scheduled trigger starts by time and does not wait for all three selected tables to update.

Use an all-table update trigger for the three source tables.

10. Add an If/else task that compares reconciliation_status: How should the workflow be configured?

1. Add a file-arrival trigger for the status.

   A file-arrival trigger detects new files and does not compare a task value.
2. Add ordinary dependencies from validation to both settlement and alerting tasks.

   Ordinary dependencies express execution relationships but do not evaluate reconciliation_status to select one branch.
3. Add an If/else task that compares reconciliation_status with "PASS". ✓

   The If/else task evaluates the task value and routes PASS to settlement and FAIL to alerting.
4. Schedule settlement during expected successful periods.

   A schedule cannot inspect reconciliation_status or reliably select a branch for each run.

Use an If/else task to compare reconciliation_status.

11. Add the notebook as a successful prerequisite: What configuration should the engineer use?

1. Give the dashboard task its own schedule shortly after the notebook schedule.

   Separate schedules provide timing but do not guarantee the dashboard waits for successful notebook completion.
2. Put both operations in one notebook and run that notebook on a schedule.

   Combining operations can impose order, but it removes the requested separate dashboard task dependency in the job DAG.
3. Use a file-arrival trigger on the dashboard's cloud-storage location.

   A file-arrival trigger does not express that the staffing notebook completed successfully before dashboard refresh.
4. Add the notebook as a successful prerequisite for the dashboard task in the job DAG. ✓

   A success dependency makes the dashboard task wait for the notebook and skip it when the prerequisite fails.

Make the dashboard task depend on successful completion of the workforce-summary notebook.

12. Add load_locations as a dependency of inventory_report: What prerequisite is missing?

1. Schedule the report after both loaders usually finish.

   A schedule cannot guarantee successful completion of both upstream tasks for each run.
2. Add load_locations as a dependency of inventory_report. ✓

   The report then has both load_inventory and load_locations as successful prerequisites.
3. Use an If/else branch to select the upstream task that finishes first.

   Selecting one task would allow the join to run without the other required input.
4. Add a file-arrival trigger for the report output directory.

   Monitoring the output directory does not ensure that both source-loading tasks completed successfully first.

Add load_locations as a dependency of inventory_report.

Other Databricks Data Engineer Associate domains

• Data Transformation and Modeling — 180 questions →
• Data Ingestion and Loading — 172 questions →
• Governance and Security — 123 questions →
• Implementing CI/CD — 82 questions →
• Troubleshooting, Monitoring, and Optimization — 82 questions →
• All 817 Databricks Data Engineer Associate questions →
• Databricks Data Engineer Associate certification: requirements, cost and exam format →