Amazon S3 annotations: attach rich, queryable context directly to objects
Amazon S3 has launched annotations, a new feature that lets users attach rich, queryable metadata directly to their objects. This allows adding structured context like descriptions, tags, or relationships alongside the object data, which can then be searched and filtered using the S3 console or API for improved data discovery and management at scale.
Background
- Amazon S3 (Simple Storage Service) is the dominant cloud object storage platform, used by millions of applications to store files, backups, data lakes, and more.
- Previously, S3 supported only "tags" (key-value pairs with a ~10-value limit per object) and "metadata" (unqueryable HTTP headers). Neither could be efficiently searched or filtered at scale.
- This new feature lets you attach structured, typed metadata (strings, numbers, dates, arrays, booleans) directly to each object and query across millions of objects using Amazon's "S3 Select" or Athena without scanning the whole bucket.
- The big shift: you no longer need a separate metadata database (e.g., DynamoDB, custom catalog) to make your S3 objects searchable — you can embed structured context alongside the data itself and run SQL-like queries on it.