A Glue job reads JSON event files in which one field sometimes holds a string and sometimes a number, which breaks a fixed schema. Which Glue abstraction handles this per-record variation without failing the job?
Show answer and explanation
Answer: C
A Glue DynamicFrame represents each record with its own schema, so a field that appears as both a string and a number can be read without a fixed upfront schema. The resolveChoice transform then casts or splits the conflicting values into consistent columns. Crawlers only update catalog metadata on their schedule and do not change how the job reads records. Catalog version history records past table definitions, and the Studio editor is a design-time tool.
Why the other options are wrong
- A. A crawler refreshes catalog metadata on a schedule but does not alter how an ETL job interprets individual records at run time.
- B. Catalog version history keeps earlier table definitions for reference, but it does not reconcile conflicting types while a job is reading data.
- D. The Studio schema editor is used to define or adjust a schema during authoring, and it does not handle per-record type variation during execution.