How is object storage different from a server's disk?

A disk (block storage) is attached to one server, limited in size, and supports in-place reads and writes — right for operating systems and databases. Object storage is unattached, effectively unlimited and HTTP-addressed, but objects are handled whole. Most real architectures use both, each for what it is shaped for.

Why does everyone implement the S3 API?

Amazon S3 arrived first and its API became the interface applications were written against. Compatible implementations let providers plug into that ecosystem instantly — for you it means tooling portability, and switching storage providers by changing an endpoint rather than your code.

What is Object Storage? — Getting Digital glossary

Object Storage

Also: S3-compatible storage, blob storage, bucket storage

Object storage keeps files as self-contained objects in flat buckets addressed over HTTP — no file system, no attached disk — scaling to practically unlimited size and billed by what you store and transfer.

The cloud's filing system isn't a file system

Disks impose limits: a volume has a size, belongs to one server, and fails with it. Object storage discards the disk metaphor. Each file becomes an object — data plus metadata under a unique key — living in a bucket that no single machine owns; the provider replicates objects across hardware behind the scenes. You talk to it over HTTP from anywhere, which is why Amazon's S3 API became the de-facto standard nearly every provider now speaks.

The trade shaping its use: objects are written and read whole. There is no seeking into a file, no locking, no directories — just keys with an illusion of folders. That makes object storage wrong for databases and live application state, and ideal for everything write-once-read-many: images and video, user uploads, backups, log archives, and static websites served straight from a bucket, usually with a CDN in front. Tiered pricing (hot, infrequent, archive) turns 'keep everything, cheaply' into an actual strategy.

Egress is where the bill lives

Storing data is cheap; moving it out is not. Object-storage pricing charges for outbound transfer, so an architecture that streams large files to users directly from a bucket can cost multiples of the storage itself. This is precisely the cost a CDN or a flat-egress provider is deployed to contain.

Key takeaways

→Objects in buckets over HTTP — no disks, no file system, no size ceiling — with the S3 API as the industry's common tongue.
→Whole-object reads and writes make it perfect for uploads, media, backups and static sites, and unsuitable for databases; watch egress, not storage, on the bill.

Related concepts

Broader topicCloud Hosting
Object storage is the cloud's native storage primitive.

FAQ

How is object storage different from a server's disk?: A disk (block storage) is attached to one server, limited in size, and supports in-place reads and writes — right for operating systems and databases. Object storage is unattached, effectively unlimited and HTTP-addressed, but objects are handled whole. Most real architectures use both, each for what it is shaped for.
Why does everyone implement the S3 API?: Amazon S3 arrived first and its API became the interface applications were written against. Compatible implementations let providers plug into that ecosystem instantly — for you it means tooling portability, and switching storage providers by changing an endpoint rather than your code.