Q2: Least-Overhead Way to Run Ad-Hoc SQL on JSON Logs in S3

Problem Info

Domain	Design High-Performing Architectures
Services	Amazon Athena, Amazon S3
Difficulty	Easy

Question

A company needs to analyze the log files of its proprietary application. The logs are stored in JSON format in an Amazon S3 bucket. Queries will be simple and run on-demand. A solutions architect must perform the analysis with minimal changes to the existing architecture and the least operational overhead.

What should the solutions architect do?

Correct Answer

Use Amazon Athena directly against the S3 bucket to run the queries as needed.

Why this is correct

The two loudest signals in this question are “on-demand / simple queries” and “least operational overhead.” That combination almost always points to Amazon Athena on the exam.

Athena is a serverless, interactive query service that runs standard SQL directly against data sitting in S3 — including JSON. There’s no cluster to provision, patch, or scale, and no data movement required: you point Athena at the bucket (optionally defining a table via the Glue Data Catalog or an inline schema), and you’re querying within minutes. You pay only for the data scanned per query, which fits “on-demand” perfectly — there’s no always-on infrastructure sitting idle between queries.

Because the logs stay exactly where they are in S3, this also satisfies “minimal changes to the existing architecture.”

Why the alternatives fall short

• Amazon Redshift / Redshift Spectrum — powerful, but provisioning a data warehouse cluster (even Spectrum needs a Redshift cluster) is significant operational overhead for “simple, on-demand” queries. Redshift shines for complex, high-concurrency BI workloads, not ad-hoc log checks.
• Amazon EMR with Hive/Presto — requires standing up and managing a Hadoop cluster. Far too much operational weight for occasional simple queries.
• Loading logs into a relational database (RDS) first — adds an ETL pipeline and a new piece of infrastructure to maintain, directly violating “minimal changes” and “least overhead.”

Exam Tip

Pattern-match this phrase combo instantly: “data already in S3” + “SQL” + “ad-hoc / on-demand” + “least operational overhead” → Amazon Athena. If the question instead emphasizes recurring, complex joins across many large datasets for dashboards/BI, lean toward Redshift. If it mentions custom big-data processing frameworks (Spark, Hive), think EMR.