Home / DE Associate practice test / Data Ingestion and Loading / Question

Databricks Data Engineer Associate · free practice question

An Auto Loader stream ingests JSON with an inferred schema. Upstream begins adding a new field. The team wants the pipeline to automatically pick up the new…

Exam-style question on Data Ingestion and Loading from the free DE Associate practice test, with the answer and a full explanation below.

Updated November 10, 2026

Data Ingestion and Loading

An Auto Loader stream ingests JSON with an inferred schema. Upstream begins adding a new field. The team wants the pipeline to automatically pick up the new column rather than silently dropping it, accepting that the stream will restart once to apply the change. Which cloudFiles.schemaEvolutionMode should be configured?

  1. Arescue — the stream never fails and never evolves the tracked schema; any newly appearing field is captured only inside the JSON _rescued_data column rather than being promoted to a real top-level column in the table.
  2. Bnone — newly appearing columns are silently ignored so the tracked schema never changes and the stream keeps running without restarting, which means the field added upstream is quietly dropped from the output rows.
  3. CaddNewColumns — Auto Loader fails the stream once when a new column appears, records it in the schema location, and the field is captured as a real column when the stream restarts, evolving the schema automatically.
  4. Dstrict — Auto Loader validates each incoming file, rejects any file that contains a column absent from the current schema, and quarantines those files for later review without ever altering the tracked schema itself.
  5. EfailOnNewColumns — the stream halts on the new column and stays failed until an engineer manually updates the schema or removes the offending files; it never adds the new column to the schema on its own at all.
Answer & explanation

Correct answer: C

WHY C: addNewColumns (the default when no explicit schema is provided) fails the stream once with UnknownFieldException on a new column, records it in the schema location, and resumes with the column included on the next start — exactly the 'auto pick up, one restart' behavior. WHY NOT A: rescue never evolves the schema; the new field would only ever live in _rescued_data, not as a real column. WHY NOT B: none silently drops new columns and never evolves — the opposite of the requirement. WHY NOT E: failOnNewColumns halts and requires manual schema intervention; it does not auto-add the column. WHY NOT D: strict is not a valid schemaEvolutionMode value.

Take the full DE Associate practice test free →

Would you pass today?

This question is one of 45 in the free, timed Databricks Data Engineer Associate practice test. Take it and get a section-by-section score.

Start the free test →