An ML engineer needs to process thousands of existing CSV objects and new CSV objects that are uploaded. The CSV objects are stored in a central Amazon S3 bucket and have the same number of columns. O

Question

Options

• AUse an Amazon Athena CREATE TABLE AS SELECT (CTAS) statement to create a table based
• BCreate a new S3 bucket for processed data. Set up S3 replication from the central S3 bucket to
• CCreate a new S3 bucket for processed data. Use AWS Glue for Apache Spark to create a job to
• DCreate a new S3 bucket for processed data. Use Amazon Data Firehose to transfer the data from

• A
  65% (32)
• B
  4% (2)
• C
  18% (9)
• D
  12% (6)

Explanation

Option A is correct because Amazon Athena is a serverless, interactive query service that runs SQL directly against S3 data with no infrastructure to provision or manage - using a CTAS statement lets you create a new table partitioned by transaction date, enabling efficient filtered queries without moving or transforming the raw files. Options B and C both require creating a separate S3 bucket and building an ETL pipeline (via replication rules or a Glue Spark job), which introduces additional components to configure, monitor, and maintain. Option D (Data Firehose) is designed for streaming ingestion from producers into destinations like S3 or Redshift - not for reprocessing existing S3 objects - making it a poor fit and high-effort solution for this use case. Athena's native integration with the AWS Glue Data Catalog also means new S3 uploads can be picked up automatically with partition projection or MSCK REPAIR TABLE, covering the ongoing ingestion requirement with minimal overhead.

Memory tip: When you see "query S3 data by a column with LEAST operational overhead," think Athena first - it's serverless SQL on S3, no ETL pipeline required. If you need partitioning for performance, CTAS is the one-shot way to reshape the table structure.

No community discussion yet for this question.