s3 on Alperen Doc

aws-s3

Mon, 01 Jan 0001 00:00:00 +0000

AWS S3#

Amazon S3 (Simple Storage Service) is AWS’ object storage service. Think of it as slower, but very scalable storage for files (objects) in containers (buckets). It’s relatively cheap and very reliable at huge scales.

By “slow,” I mean it’s not designed for low-latency access like a database, and it’s not designed to be easily searchable. It’s designed for storing and retrieving files by their key (filename).

aws-s3-bucket

Mon, 01 Jan 0001 00:00:00 +0000

S3 Buckets#

The bucket is the highest level of organization in S3; it’s a container for storing objects.

Some projects put everything in a single bucket, organizing objects by prefix (like folders). Other projects use multiple buckets for different purposes (for example, one for user uploads and another for static assets). There’s no single right answer; it depends on your needs.

That said, you don’t want to find yourself searching across multiple buckets for a file, so try to keep any collection of related files in the same bucket.

aws-s3-objects

Mon, 01 Jan 0001 00:00:00 +0000

S3 Objects#

You can think of “object storage” and “file storage” as almost the same thing, and “objects” as “files.” The big technical difference is that objects are stored in a way that allows more efficient distribution across many machines (which is why S3 is so scalable and durable), but from a practical user’s standpoint, they’re very similar.

Storing stuff in buckets is simple: file A goes in bucket B at key C. You only need two things to access an object in S3: