Athena

• Is serverless to query S3 using SQL
• Uses SQL language to query files (Build on top of Apache Presto)
• It's schema on read
• Accepts format:
  • CSV
  • Json
  • Parquet (columnar, splittable)
  • ORC (columnar, splittable)
  • Avro (columnar, splittable)
  • XML
• Princing: fixe amount per TB of data scanned
• Structured, semi-structured an unstructured data
• Uses cases: BI, analytics, reporting, query logs, Ad-hoc queries
• Athena detect automatically the tables from Glue, and it creates queryable table

Performance

• Use columnar data for cost-savings (less scan)
  • Apache Parquet or ORC is recommended
  • Use Glue to convert data to Parquet or ORC
• Compress data for smaller retrievals (bzip2, gzip, lz4, snappy, zlip, zstd, ...)
• Partition datasets in S3 for easy querying on virtual columns
• Use larger files ( > 128 MB) to minimize overhead

Costs

• Pay-as-you-go
  • $5 per TB scanned
    • Successful or cancelled queries count, failed queries do not
    • No charge for DDL (Create/ Alter, Drop, ...)
  • Save money by using columnar formats
    • ORC, Parquet
    • Save 30-90% and get better performance
    • Good partitioning (Less query CPU)
  • Glue and S3 have their own charges

Federated Query

• Query data anywhere (S3 ou on-premise)
• Use Data Source Connectors thar run on AWS Lambda to run Federated Queries

Security

• Access control
  • IAM, ACLs, S3 bucket policies
  • AmazonAthenFullAcces/
• Encrypt .... ???
• Cross-acount in S3 bucket policy possible
• Transport Layer Security (TLS) encrypts in-transit (between Athena and S3)

Anti-Pattern

• Highly formatted reports / visualization => That's what QuickSight is for
• ETL (not suitable for ETL actions) => use Gue instead