Data Engineer Associate · 10% of the exam

Troubleshooting, Monitoring, and Optimization: free practice questions

5 sample questions from our 5-question bank for this domain — answers and explanations included. These are the same scenario-based style as the real Databricks exam.

1. Queries frequently filter a large Delta table by a high-cardinality column. Which technique co-locates related data to speed those filters?

  • A. Z-ORDER (or liquid clustering) on that column✓ Correct
  • B. Dropping the column
  • C. Running VACUUM with a 0-hour retention
  • D. Converting the table to CSV
Explanation

Z-ORDER (or liquid clustering) co-locates related values so filters skip more files (data skipping). Dropping the column (B), aggressive VACUUM (C), and CSV conversion (D) don't improve filtered reads.

2. Which command permanently removes data files no longer referenced by a Delta table (respecting the retention window) to reclaim storage?

  • A. VACUUM✓ Correct
  • B. OPTIMIZE
  • C. MERGE INTO
  • D. CREATE TABLE
Explanation

VACUUM deletes files no longer referenced by the table beyond the retention threshold, reclaiming storage. OPTIMIZE (B) compacts, MERGE (C) upserts, and CREATE TABLE (D) makes a table.

3. A streaming ingest has produced millions of tiny files, slowing queries. Which Delta maintenance command compacts them into larger files?

  • A. OPTIMIZE✓ Correct
  • B. VACUUM
  • C. DESCRIBE HISTORY
  • D. REFRESH TABLE
Explanation

OPTIMIZE compacts many small files into fewer larger ones, improving read performance. VACUUM (B) removes old unreferenced files, DESCRIBE HISTORY (C) shows the commit log, and REFRESH (D) reloads metadata.

4. Where can a data engineer see how long each task took and read the error/output logs for a job run?

  • A. The job run details and task logs in the Jobs UI✓ Correct
  • B. The Delta transaction log only
  • C. A random notebook cell
  • D. The cluster's billing invoice
Explanation

The Jobs UI shows per-task durations, statuses, and logs for each run, which is where you diagnose failures. The Delta log (B) tracks table commits, and options C/D don't hold run logs.

5. A scheduled job failed on its third task due to a transient error. What is a reasonable first step to recover without rerunning successful tasks?

  • A. Use 'Repair run' to rerun only the failed (and downstream) tasks✓ Correct
  • B. Delete the entire job
  • C. Rewrite all the code from scratch
  • D. Disable Unity Catalog
Explanation

Repairing a run reruns only the failed and dependent tasks, reusing successful results. Deleting the job (B), rewriting everything (C), and disabling UC (D) are not appropriate first steps.

0 more questions in this domain

Practice the full bank with instant grading, flashcards, and a timed mock exam.

Start practicing free
Troubleshooting, Monitoring, and Optimization — Free Data Engineer Associate Practice Questions | DataCertPrep — Certification Prep