S3 Bucket

An S3 bucket is file storage — a durable place to keep files your app receives or produces: profile pictures, document uploads, generated PDFs, exports, backups. Files put in S3 are stored redundantly and stay available for years.

You use S3 instead of saving files on the server itself, because servers come and go on every deploy — anything written to them is lost. S3 outlives deploys.

When you need it

• Users upload files (avatars, images, attachments).
• Your app generates files to download later (invoices, reports, exports).
• You need somewhere to keep backups or large assets.

How it connects

Link the bucket to a service and StrictOps injects the bucket name:

S3_BUCKET=your-bucket-name

Your app uses the AWS SDK with this bucket name to upload and download objects. Access is scoped to the linked services, so only they can read and write the bucket.

Settings explained

• Region — which AWS region the bucket lives in. Keep it the same as your services for speed and lower transfer cost.
• Object versioning — keeps previous versions of a file when it's overwritten, so you can recover from accidental changes or deletes. Handy for important documents; off by default.
• Public read access — makes objects readable by anyone with the URL. Only enable this for genuinely public assets (for example, images on a marketing page). Leave it off for anything user-private.

Cost guidance

S3 is inexpensive and you pay for what you store and transfer, so an empty bucket costs almost nothing. Versioning increases storage over time because old copies are retained.