Object Storage

Object storage is a way of storing data as self-contained objects, each consisting of the data itself, rich metadata, and a globally unique identifier. Objects live in a flat namespace called a bucket or container and are accessed over HTTP APIs rather than a file-system path or block device. Amazon S3, Azure Blob Storage, and Google Cloud Storage are leading examples.

How It Works

Instead of organizing data into a hierarchy of folders or fixed-size blocks, object storage gives each object a key and stores it in a bucket. Clients create, read, and delete objects through REST APIs. The system handles replication and durability automatically, often spreading copies across availability zones to achieve very high durability. Features include versioning, lifecycle policies that move or expire objects, server-side encryption, fine-grained access policies, and event notifications. Object storage is not designed for frequent in-place edits; objects are typically written and read whole.

Why It Matters

Object storage is the default home for unstructured data at scale: backups, logs, media, data-lake datasets, static website assets, and machine-learning training data. It is cheap, effectively unlimited in capacity, and highly durable, with tiered pricing that lets cold data move to lower-cost classes. Its API-based access and metadata make it a natural integration point for data pipelines and analytics. The trade-offs are higher latency than block storage and no support for partial updates, so it is unsuitable as a database disk or for transactional workloads.

Related Terms

Object storage contrasts with block storage, builds on infrastructure-as-a-service, and commonly feeds data pipelines and content delivery networks.