Original questions, written to the 2026 exam guide
Why these and not a dump
These are practice questions, not leaked exam content. Each one teaches the concept the real exam tests, with an explanation you can check against the Databricks docs. Why dumps fail →
Databricks Intelligence Platform
- A regulatory audit requires the team to reproduce exactly what a Delta table looked like at a specific timestamp two weeks ago, and to roll…
- Multiple analysts run frequent ad-hoc SQL against curated Delta tables throughout the day. Requirements: fast startup, high concurrency for…
- A data engineer is deciding which of several workloads to move onto Photon-enabled compute to gain the most speedup. They need to know…
Data Ingestion and Loading
- A data engineer must incrementally load CSV files that land a few times per day in an S3 path into a Unity Catalog managed table. The load…
- A team must ingest CRM records from Salesforce into Unity Catalog-governed tables on a reliable schedule, with minimal custom code and…
- A data engineer sets up an Auto Loader stream with spark.readStream.format('cloudFiles') to ingest JSON files as they land in cloud…
- An engineer wants to quickly preview a few CSV files sitting in a cloud storage path from the SQL editor, just to inspect their contents…
- An Auto Loader stream ingests from a bucket that accumulates millions of files. Directory-listing mode has become slow and expensive…
- A data engineer must ingest data from a SaaS Salesforce source and a self-managed SQL Server database into Unity Catalog tables using…
- Incoming JSON events contain a nested `address` object with `city` and `zip` fields. A silver table needs `city` and `zip` as top-level…
- An Auto Loader stream ingests JSON with an inferred schema. Upstream begins adding a new field. The team wants the pipeline to…
- A data engineer wants an Auto Loader pipeline to run on a nightly schedule: each run should process every file accumulated since the…
Data Transformation and Modeling
- A data engineer receives a daily batch of customer records containing both brand-new customers and updates to existing ones, keyed by…
- In a Lakeflow Spark Declarative Pipeline, a bronze streaming table ingests append-only clickstream events. A downstream Gold object must…
- A data engineer adds a data-quality constraint to a Lakeflow declarative pipeline table so that any record failing the constraint is…
- A data engineer sees that a Delta table has accumulated many small files from frequent streaming writes, slowing queries that filter on a…
- A data engineer needs downstream consumers to read row-level inserts, updates, and deletes from a Delta table incrementally, without…
- A dashboard needs a fast, approximate count of unique visitors over billions of rows, where a small error margin is acceptable in exchange…
- A silver transformation must drop any row where the `user_id` column is null, while keeping rows that have nulls in other columns. Which…
- A Lakeflow declarative pipeline consumes a CDC feed of inserts, updates, and deletes keyed by customer_id, and events sometimes arrive out…
- A transformation must return all rows from a `orders` DataFrame and attach matching customer attributes from a `customers` DataFrame…
- A bronze table may contain duplicate event rows that are identical across all columns due to at-least-once delivery. The silver job must…
Governance and Security
- A security team wants to ensure that one specific user can never read main.hr.salaries, even though that user belongs to a group that has…
- A data engineer needs to let a team read and write data files at a specific cloud storage path that is registered in Unity Catalog as an…
- A team creates a Unity Catalog managed table and later runs `DROP TABLE`. A separate external table points at an existing cloud storage…
- A team must give a partner organization live, read-only access to a Delta table. The partner does not use Databricks and has no account in…
- An analyst reports they cannot query main.sales.orders even though the table owner ran GRANT SELECT ON TABLE main.sales.orders to the…
- A table contains a `ssn` column that must be visible in full only to the `pii_admins` group; everyone else should see it masked. The…
- A data engineer wants to reference a table in a query using its fully qualified name in Unity Catalog. The table is named orders, it lives…
Implementing CI/CD
- In a CI pipeline running on a pull request, the team wants to confirm an Automation Bundle's configuration is syntactically valid and all…
- A data engineer is initializing a new Databricks Asset Bundle project and must create the single required root configuration file that…
- A team's PySpark transformation logic is factored into pure functions. On every pull request, a CI runner must execute a pytest suite…
- A data engineer wants to develop pipeline code in an isolated branch inside the Databricks workspace, commit changes, and open a pull…
- A single Automation Bundle (formerly Databricks Asset Bundles) must deploy the same job to dev, staging, and prod, each using a different…
Troubleshooting, Monitoring, and Optimization
- A stage runs far slower than expected. In the Spark UI stage details the data engineer sees large non-zero values under the Spill (memory)…
- In the Spark UI a data engineer sees one join stage where the max shuffle-read task is 5 GB while the median is 60 MB, and a single…
- A data engineer joins a large fact table to a small 8 MB dimension table but notices Spark ran a shuffle sort-merge join instead of a…
- A Delta table fed by many small streaming micro-batches has accumulated thousands of tiny files, and interactive reads over it have grown…
Working with Lakeflow Jobs
- A team runs a nightly scheduled Lakeflow Job in production. It is currently attached to a shared all-purpose cluster that several analysts…
- In a Lakeflow Job, a notification task must run after three upstream tasks regardless of whether they succeed or fail, so alerts are always…
- A data engineer builds a Lakeflow Job with four tasks: ingest_bronze, clean_silver, aggregate_gold, and refresh_dashboard. clean_silver…
- A data engineer schedules a Lakeflow Job to run every 15 minutes. Occasionally a run takes longer than 15 minutes. The engineer notices…
- A task occasionally fails due to a transient network timeout when calling an external API, but succeeds when re-run. Without adding…
- A data engineer configures a Lakeflow Job with a file arrival trigger on an external location where an upstream system uploads a batch of…
- A downstream reporting job must run only after new files actually arrive in a landing zone, which happens irregularly — sometimes several…
Score it like the real exam
The same 45 questions, timed, with a section-by-section breakdown at the end. Free, no credit card.
Start the free DE Associate practice test →