Object Storage

Object storage stores files as objects inside buckets, accessed through an HTTP API (usually S3-compatible). There is no hierarchical filesystem like regular folders; each object has a key, metadata, and URL.

Hosting use cases

How it differs from block storage

Block storage is a disk attached to one server and formatted with a filesystem. Object storage is accessed over HTTP and designed for large scale. For applications that need ordinary read/write file access, block storage is the better fit.

What to watch for

• Egress cost. Downloading data can cost more than storing it.
• Consistency. Some services have specific consistency behavior; read the documentation.
• Access policy. Public buckets can leak; use private buckets and signed URLs.
• Versioning. Enable it when you need protection from overwrites or accidental deletes.
• Never put credentials in the client. Access through the backend or temporary tokens.
• Put a CDN in front. For public assets, a CDN reduces latency and direct egress load.

To move WordPress media to object storage, use an S3-compatible plugin and configure public URLs or a CDN. Test uploads, deletes, and thumbnail regeneration before production. Do not move theme or plugin files to object storage; they stay in the document root. Use versioning for backup buckets and signed URLs for private files.

An upload example with the AWS CLI:

aws s3 cp ./image.jpg s3://my-bucket/media/image.jpg --acl private

Use --acl private and signed URLs for files that must not be public. For public assets, put a CDN in front of the bucket and set cache headers on the object.

FAQ