Amazon Athena

A service to query logs and CSVs stored in S3 with SQL.

Back to glossary Go to related practice questions

In one sentence

A service to query logs and CSVs stored in S3 with SQL.

Overview

Amazon Athena is a serverless query service that analyzes data on S3 using SQL.

Common use cases

S3 log analysis
Lightweight analysis of CSV or Parquet files
Ad-hoc queries on a data lake

CLF-C02 exam points

A serverless analytics service
Queries data on S3 using SQL
Billed based on the amount of data scanned

SAA-C03 exam points

Using formats like Parquet reduces the amount of data scanned
Combine with Glue Data Catalog to manage table definitions
A good candidate for analyzing CloudFront or S3 logs

Practical Use

Amazon Athena is a serverless analytics service that lets you run standard SQL queries against data stored in S3. There is no need to provision a database server in advance — just place data in S3, register table definitions in Glue Data Catalog or similar, and queries can be executed immediately. In practice, Athena is used for analyzing CloudFront and ALB access logs, auditing CloudTrail logs, performing ad-hoc investigation on business data accumulated in CSV or JSON, and exploring data in a data lake. Athena pricing is based on the amount of S3 data scanned per query (a few USD per TB). Therefore, the design of data formats and partitioning is the central concern for cost optimization. Converting row-oriented formats such as CSV or JSON to columnar formats such as Parquet or ORC significantly reduces scan volume because only the necessary columns are read. Additionally, partitioning by frequently filtered keys such as date or region allows Athena to automatically skip scanning unnecessary partitions. A standard design pattern is to manage table definitions (schema and partition information) centrally in Glue Data Catalog and share them across Athena and other analytics services (Redshift Spectrum, EMR, etc.). CLF-C02 / SAA-C03 ask about Athena as "a serverless service that allows SQL analysis of data in S3". Differentiation from EMR (managed Hadoop/Spark) and Redshift (data warehouse) is also a frequent topic. The standard division is: Athena for ad-hoc queries and small to medium-sized log analysis, EMR for large-scale batch processing, and Redshift for routine BI analysis. Remembering this segmentation makes judgment calls easier.

Official Documentation

Cost notes

Charged based on the volume of data scanned by queries
Avoid unnecessary full scans

Security notes

Separate S3 bucket permissions from Athena execution permissions
Manage permissions on the S3 location used for query results

Related services

Amazon Simple Storage Service

A representative AWS storage service for storing files securely and at low cost.

AWS Glue

Service for organizing analytical data and converting it into usable forms.

Amazon CloudFront

A CDN that delivers web content quickly and securely from locations close to the user.

Related architecture diagrams

Related architecture diagrams have not been added for this term yet.