Data-Engineer-Associate - Governance and Security (15% of the exam) - Section 7.1

Differentiate between managed and external tables in Unity Catalog and perform basic operations on them, including converting between the two.

Distinguish managed from external tables by who controls the storage and what DROP TABLE does to the data. Create, alter, drop and convert between the two, and recognise the role of external locations and storage credentials.

managed tablesexternal tablesexternal locationsstorage credentialsDROP TABLE behaviour

Practice question for this objective

Free sampleGovernance and Securitymedium

A team owns an external Delta table registered over a path inside a working external location. No system outside Databricks reads that path any more, so the team wants Unity Catalog to own the table's data lifecycle and its file maintenance from now on, while keeping the same three level name and the table's existing history. Which approach converts the table in place?

ARun CONVERT TO DELTA on the table, which rewrites the transaction log so that Unity Catalog takes ownership of the files stored under the external path.
BDrop the external location that governs the path, which leaves the table registered with no governed path, so Unity Catalog reclassifies it as a managed table on the next write.
CSet the table property that marks it as managed with ALTER TABLE SET TBLPROPERTIES, which tells Unity Catalog to begin managing the files already present under the external path.
DRun ALTER TABLE with SET MANAGED, which converts the external table to a managed table under the same name while preserving its data and history. Correct

ALTER TABLE with SET MANAGED converts an external Unity Catalog table into a managed table in place. Converting an external table to a managed table hands the data lifecycle to Unity Catalog, which then owns file placement, deletion and automatic maintenance. Databricks provides this as an in place operation through ALTER TABLE with SET MANAGED, so the table keeps its catalog, schema and name and its history, rather than requiring a copy into a new table and a rename.

Why A is wrong: Tempting because the command name suggests a conversion, but CONVERT TO DELTA turns a Parquet or Iceberg dataset into a Delta table and has no effect on whether a table is managed or external.

Why B is wrong: Dropping the external location removes the authorisation path rather than changing the table type, so queries against the external table would begin failing instead of the table becoming managed.

Why C is wrong: Plausible because many Delta behaviours are controlled by table properties, but table type is not a Delta table property and cannot be flipped with SET TBLPROPERTIES.

Why D is correct: Correct, because SET MANAGED is the supported in place conversion from an external table to a managed table, keeping the identifier, the data and the table history intact.

See more Data-Engineer-Associate practice questions, answers explained.

Exam traps in Governance and Security

Answers that look right on this material and are not. Each one is a distractor from a different question in the Data-Engineer-Associate bank for this domain.

Both tables write into the schema's managed storage location, and the LOCATION clause on orders_b only records a path that Unity Catalog reports in table metadata.
Why it is wrong: It is tempting because the schema's managed storage location is real and does govern orders_a, but LOCATION is not merely descriptive: it makes orders_b an external table whose files are written to the supplied path.
An external table, because only an external table can be secured with Unity Catalog privileges and have its reads and writes captured in the metastore audit trail alongside the team's other objects.
Why it is wrong: Governance is a real reason to register a dataset at all, but it does not separate the two types: managed and external tables are both secured by Unity Catalog privileges and both appear in the audit trail.
Run ALTER TABLE with SET LOCATION against the managed table, which moves its files to the supplied cloud path and reclassifies it as an external table under the same name.
Why it is wrong: Tempting because the clause names a path, but it does not relocate a managed table's data or turn it into an external table, so the partner tool would find nothing at the path.

Examworthy is not affiliated with or endorsed by Databricks. Original, blueprint-aligned practice material only.