A daily Glue ETL job reprocesses the entire S3 source every run even though only new partitions change. Which feature makes the job process only new data?
A.Glue job bookmarks
Correct. Bookmarks = incremental processing state.
B.Larger worker types
Faster full reprocessing is still full reprocessing.
C.Glue crawlers on a schedule
Crawlers update the catalog — they don't filter source data.
D.Enabling job metrics
Metrics observe the job, not change its inputs.
Why this answer
Job bookmarks persist state between runs so Glue tracks what was already processed and handles only new/changed data — the built-in incremental-ETL mechanism. Bigger workers speed the same wasted work; crawlers discover schema, not processed state.
Correct answer: A · Last reviewed 2026-11-02
Official AWS references
Practice this interactively
Answer without seeing the solution first in the Data Operations practice session.
Practice Data Operations