Object storage

Object storage is a way to store data as discrete objects — the file bytes plus metadata and a unique key — in a flat namespace instead of a file-system hierarchy. It scales to billions of objects, is accessed over HTTP APIs, and underpins most cloud file storage, backups, and static assets.

Unlike a traditional file system with nested directories, or a block device with raw volumes, object storage treats each file as a self-contained object addressed by a key. There are no real folders — the slashes in a key are just part of its name, which is why object storage scales to enormous object counts without the overhead of a directory tree.

Each object carries metadata alongside its bytes: content type, cache directives, custom headers, and an ETag. You interact with it over HTTP — PUT to write, GET to read, DELETE to remove — which makes it a natural fit for web apps, APIs, and direct browser uploads.

The trade-off is that objects are typically immutable: you replace a whole object rather than editing it in place, and you don't get file-locking or partial random writes the way a POSIX file system offers. That constraint is what buys the durability and scale.

Related terms

Bucket

A bucket is a top-level container that holds objects in cloud storage, identified by a globally unique name. It's an Amazon S3 term — since adopted by many S3-compatible services — that sets a namespace and boundary for permissions, region, and configuration. Every stored object belongs to exactly one bucket.

Blob storage

Blob storage holds binary large objects — images, video, backups, or any unstructured file — as opaque chunks of bytes accessed by key over HTTP. "Blob" is Microsoft Azure's name for what other providers call object storage. It's built for massive scale and direct delivery, not relational queries.

Multipart upload

Multipart upload splits a large file into smaller parts that are uploaded independently — often in parallel — then reassembled into one object on the server. It improves throughput, survives flaky networks by retrying only failed parts, and is the standard way to move large files into object storage reliably.

ETag

An ETag (entity tag) is an identifier a server assigns to a specific version of a resource, often a hash of its contents. Clients send it back via the If-None-Match header so the server can reply 304 Not Modified when nothing has changed, saving bandwidth. It also helps detect mid-update conflicts.

How files.link fits in

files.link gives you object-storage scale and HTTP delivery without provisioning or tuning any of the underlying storage yourself. a simpler S3 alternative.

Get started

← Back to the glossary