To minimize processing overhead per file, what should you do with many small data files?

• A. Aggregate them into larger files
• B. Merge them into a single file
• C. Split them into individual files
• D. Load them as-is with no changes

Answer: (A)

When processing data, there’s overhead tied to each file you touch—opening it, reading its metadata, and scheduling work for it. Having many tiny files means lots of these per-file costs add up, which can throttle throughput and waste time on metadata lookups and task setup. By aggregating those small files into fewer, larger files, you cut down the number of file handles the system must manage, reduce metadata operations, and improve I/O efficiency. This generally yields faster reads and better overall processing performance. (Just be mindful not to createfiles so large that they hinder parallelism or become unwieldy.)