Home / DE Associate practice test / Data Governance & Quality

Free · 3 questions with explanations

Data Governance & Quality: Databricks Data Engineer Associate Practice Questions

Exam-style questions on Data Governance & Quality. Pick your answer, then open the explanation to see why it's right — and why the other options are wrong.

Updated November 10, 2026

1 Data Governance & Quality

A data engineering team is migrating a legacy Hive metastore data warehouse to Unity Catalog. They have two types of data assets: - **Dataset A**: Internal Silver-layer Delta tables produced entirely by Databricks pipelines. No external tool reads these tables directly from cloud storage. - **Dataset B**: Raw CSV and Parquet landing files produced by an upstream on-premises ETL system that writes directly to an S3 bucket. Databricks only reads this data. A data engineer proposes registering Dataset B as a managed table in Unity Catalog by running CREATE TABLE ... AS SELECT ... to import the data. A colleague argues this would be the wrong table type for Dataset B and instead recommends an external table. Which statement BEST explains the colleague's reasoning, and what is the KEY behavioral difference between a managed and external table in Unity Catalog when a DROP TABLE command is executed?

  1. AThe colleague is correct. Unity Catalog fully manages a managed table's lifecycle: DROP TABLE deletes its metadata and, after a recovery period (7 days by default), its data files. For Dataset B, whose files an upstream system owns in S3, an external table fits: DROP TABLE only removes the metadata and leaves the files untouched.
  2. BThe colleague is correct. An external table in Unity Catalog enables full read and write access from both Databricks and any external cloud storage tool simultaneously, allowing the on-premises ETL system to append new CSV files while Databricks reads them concurrently without any consistency issues. By contrast, a managed table exclusively locks the storage path, preventing the upstream on-premises system from writing new files to the same S3 bucket location while the table is registered in Unity Catalog.
  3. CThe colleague is incorrect. Managed tables and external tables are functionally identical after registration in Unity Catalog—both types support the same file formats (Delta, CSV, Parquet, Avro), and DROP TABLE behaves identically for both, removing only the metadata registration while always preserving the underlying physical data files to prevent accidental data loss in a shared environment.
  4. DThe colleague is incorrect. Since Dataset B is read-only from Databricks, registering it as a managed table is the right choice because managed tables in Unity Catalog offer superior query performance through predictive optimization and automatic clustering, regardless of whether the underlying data was created by Databricks. The DROP TABLE behavior is irrelevant for read-only datasets because Unity Catalog enforces a data protection policy that prevents DROP TABLE from deleting files when the table has been read by two or more users in the past 30 days.
Show answer & explanation

Correct answer: A

WHY A: This is the definitive distinction between managed and external tables. Unity Catalog managed tables use Delta format and UC owns the full data lifecycle—DROP TABLE deletes both metadata AND physical files (the files after a recovery period, 7 days by default, during which UNDROP TABLE can restore the table). For Dataset B (upstream system writes raw files to S3), an external table is correct: DROP TABLE only de-registers the metadata, leaving the physical files safe for the upstream system's continued use. WHY NOT B: Unity Catalog does not 'lock' a managed table's storage path to prevent external writes—it simply doesn't govern those external writes. Concurrent external writes to a managed table's storage location are possible but bypass UC governance, which is why they're discouraged. WHY NOT C: Managed tables MUST use Delta (or Apache Iceberg in Public Preview) format. External tables support Delta, CSV, JSON, Avro, Parquet, ORC, and Text. DROP TABLE behavior is fundamentally different: managed deletes data files, external preserves them. WHY NOT D: Performance advantages of managed tables are real, but the DROP TABLE behavior claim at the end is entirely fabricated—there is no such 30-day read-based data protection policy in Unity Catalog.

2 Data Governance & Quality

A data engineering team is evaluating whether to use **Lakehouse Federation** or **managed ingestion (Auto Loader / COPY INTO)** for connecting to several external data sources. They describe the following four scenarios: - **Scenario A**: A data scientist wants to run exploratory SQL queries against a production PostgreSQL database to understand data shape before designing an ETL pipeline. Speed of setup is more important than query performance. - **Scenario B**: A critical nightly pipeline must join 500 million records from an internal Delta table with 200 million records from a Snowflake data warehouse. The result feeds a real-time executive dashboard that requires sub-3-second query times. - **Scenario C**: A compliance team needs to query a MySQL database that stores audit records from a legacy HR system. They want the data governed by Unity Catalog (lineage, fine-grained access control) without moving data into Databricks. - **Scenario D**: A streaming pipeline must ingest IoT sensor data arriving at 100,000 events per second from an external Kafka topic and land it into a Unity Catalog Delta table for downstream real-time analytics. For which scenarios is **Lakehouse Federation** the MOST appropriate solution?

  1. ALakehouse Federation is most appropriate for Scenarios A and B. Scenario A fits: federation gives fast setup for ad-hoc exploration of PostgreSQL with no ingestion step. Scenario B fits: with query pushdown, the 500M × 200M join is planned by Lakehouse Federation and executed inside Snowflake's compute layer, which removes the data transfer and delivers the sub-3-second dashboard queries with minimal Databricks compute.
  2. BLakehouse Federation is most appropriate for Scenarios B and D. Scenario B fits: federation supports high-throughput joins between Delta tables and external warehouse tables, and its caching layer materializes the Snowflake results for dashboard serving. Scenario D fits: a foreign catalog can register a Kafka topic as a federated streaming source, with Unity Catalog governing the stream's schema and access policies.
  3. CLakehouse Federation is most appropriate for Scenarios A and C. Scenario A (ad-hoc exploration of PostgreSQL ahead of an ETL design) fits: federation queries the source in place with no data migration, giving fast setup for proof-of-concept work. Scenario C (a MySQL audit database queried without moving data) fits: a foreign catalog extends Unity Catalog governance, such as lineage and fine-grained access control, to the external source.
  4. DLakehouse Federation is most appropriate for Scenarios C and D. Scenario C fits: foreign catalogs place the external MySQL data under Unity Catalog governance with no data movement. Scenario D fits: a streaming foreign catalog reads the Kafka topic natively inside Unity Catalog and applies Delta Lake ACID guarantees to the incoming IoT events, with no separate Auto Loader or Structured Streaming configuration.
Show answer & explanation

Correct answer: C

WHY C: Lakehouse Federation is specifically recommended for (1) ad-hoc exploration and proof-of-concept work where you want to query external data without ingestion (Scenario A, PostgreSQL exploration), and (2) scenarios where you need Unity Catalog governance (lineage, fine-grained access control) applied to external sources without data migration (Scenario C, MySQL HR audit data). Both match the official Databricks recommended use cases for Lakehouse Federation. WHY NOT A: Scenario A is right for Lakehouse Federation, but Scenario B is NOT. Sub-3-second joins of 500M × 200M records require the data to be ingested into Databricks (or cached). Lakehouse Federation does support query pushdown and caching, but a join against an internal Delta table cannot run wholly inside Snowflake, and for high-volume, low-latency production joins feeding real-time dashboards, managed ingestion into Delta is the recommended approach. WHY NOT B: Scenario B's latency requirement (sub-3-second) makes Lakehouse Federation unsuitable for the join itself. Scenario D (100K events/second from Kafka) is a streaming ingestion use case; Lakehouse Federation does NOT support Kafka as a foreign catalog source and is designed for SQL database federation (MySQL, PostgreSQL, Snowflake, etc.). Use Structured Streaming for Kafka. WHY NOT D: Scenario C is right, but Scenario D is incorrect for Lakehouse Federation. There is no 'streaming foreign catalog' feature for Kafka in Lakehouse Federation; this is a fabricated capability.

3 Data Governance & Quality

A data platform team at a retail company needs to share sales analytics data with three different recipients: - **Recipient 1**: An internal data science team at the same company using a different Databricks workspace in a different cloud region. - **Recipient 2**: A strategic retail analytics partner that uses Apache Spark (open source, not Databricks) to process shared data in their own environment. - **Recipient 3**: A business intelligence team at a subsidiary company using Tableau connected to their own Databricks SQL warehouse. After reviewing Delta Sharing documentation, a data engineer lists the following statements about Delta Sharing capabilities and limitations. Which statement is ACCURATE?

  1. ADelta Sharing gives recipients read/write permissions on shared data, and Recipient 1's data science team can commit model output directly into the provider's shared tables. This two-way access is the main advantage of Databricks-to-Databricks Delta Sharing over open sharing, which supports one-directional, read-only access for non-Databricks consumers such as Recipient 2; Recipient 3 would likewise be able to write back from its Databricks SQL warehouse.
  2. BDelta Sharing is designed for data exchange between separate organizations and cloud providers and does not cover internal cross-workspace or cross-region sharing within the same Databricks account. For Recipient 1 (same company, different workspace and region), the appropriate approach is an external table reachable from both workspaces through a common external location, with Delta Sharing reserved for Recipient 2 and Recipient 3, which sit outside the company's primary metastore.
  3. CDelta Sharing requires recipients to run on the Databricks platform, whether they are internal or external, as the protocol depends on Unity Catalog's proprietary credential vending mechanism, which is compatible with Databricks Runtime. Recipient 2, on open-source Spark, can consume the shared data once it installs the official Databricks Spark connector and authenticates with a Databricks personal access token issued by the provider organization.
  4. DFor Recipient 1 (separate cloud region), Databricks-to-Databricks Delta Sharing applies, although Unity Catalog lineage graphs do not extend across regions or platforms and will not show the cross-region flow. For Recipient 2 (open-source Spark), open Delta Sharing applies: the recipient reads the shared data as a read-only dataset through the open protocol with no Databricks account. For each recipient, Delta Sharing grants read-only access, with no path to write back to the provider's tables.
Show answer & explanation

Correct answer: D

WHY D: This answer is accurate on all three points. (1) Databricks-to-Databricks Delta Sharing works cross-region, but Unity Catalog lineage graphs do not cross regional or platform boundaries; lineage tracking stops at the metastore level and will not show cross-region data flow, per official Databricks best practices documentation. (2) Open Delta Sharing (non-Databricks recipients) allows systems like open-source Spark to consume shared data as read-only using the open Delta Sharing protocol without a Databricks account; this is a core design goal of the open-source Delta Sharing standard. (3) Delta Sharing is strictly read-only for all recipient types; no recipient can write back to the provider's shared tables through Delta Sharing, whether it is Databricks-to-Databricks or open sharing. WHY NOT A: Delta Sharing does NOT grant write access to recipients. All sharing types, including Databricks-to-Databricks, are strictly read-only. Recipients cannot write model outputs or any data back to provider tables through the sharing mechanism. WHY NOT C: Delta Sharing IS an open-source protocol specifically designed for non-Databricks consumption. External systems such as open-source Apache Spark, Pandas, PowerBI, Tableau, and others can consume Delta Sharing data using open client libraries without a Databricks account or any proprietary Databricks connector. WHY NOT B: Delta Sharing CAN be used for internal cross-workspace and cross-region data sharing within the same organization. Databricks-to-Databricks Delta Sharing is specifically the recommended approach for sharing data between workspaces in different regions of the same Databricks account.

Take the full DE Associate practice test →