A municipal transfer station lands 80 daily weigh-ticket columns on Amazon S3. Analysts later scan only five columns with Amazon Athena before the set is handed to Amazon SageMaker. Which file layout should they choose?
Select an answer to reveal the explanation.
Short Explanation
Think of 80 weigh-ticket columns sitting on S3, and Athena later reading only five. Pick Apache Parquet. A CSV dump makes every query reread the other 75, and RecordIO is a trainer wrap, not the lake.
Full Explanation
Column-subset scans through Athena favor a columnar layout such as Apache Parquet. A row-oriented CSV dump rereads unused columns. RecordIO is an ingest convention for some SageMaker built-in trainers, not the lake-analytics default. EBS is not the durable lake.