A data analyst is designing a solution to interactively query datasets with SQL using a JDBC connection. Users will join data stored in Amazon S3 in Apache ORC format with data stored in Amazon…

Question

Options

• AUse AWS Glue jobs to ETL data from Amazon ES and Aurora MySQL to Amazon S3. Query the
• BUse Amazon DMS to stream data from Amazon ES and Aurora MySQL to Amazon Redshift.
• CQuery all the datasets in place with Apache Spark SQL running on an AWS Glue developer
• DQuery all the datasets in place with Apache Presto running on Amazon EMR.

• A
  8% (3)
• B
  21% (8)
• C
  13% (5)
• D
  58% (22)

Explanation

Apache Presto running on Amazon EMR (Option D) provides the most up-to-date results because it queries all data sources in place - no data movement or replication. Presto is a distributed SQL query engine that supports federated queries across heterogeneous sources simultaneously: S3 (ORC format via the Hive connector), Amazon Elasticsearch Service (via the Elasticsearch connector), and Aurora MySQL (via the MySQL connector). It exposes a JDBC endpoint, directly satisfying the connectivity requirement. Option A uses Glue ETL jobs to copy data to S3 first, introducing staleness. Option B uses DMS to stream to Redshift, adding replication lag and extra infrastructure. Option C (Spark SQL on a Glue developer endpoint) is not designed for end-user interactive JDBC access and lacks native, easy-to-configure connectors for Elasticsearch.

No community discussion yet for this question.