What Is Object Storage? Files at Cloud Scale
Object storage stores data as objects (a file plus metadata and a unique key) in flat containers called buckets, accessed over HTTP APIs and built for massive scale and durability.
Object storage is the cloud way to store unstructured data: images, backups, logs, build artifacts, and static sites. Unlike a filesystem with nested directories or a block device for a database, it exposes a flat key-value model over HTTP. S3, Google Cloud Storage, Azure Blob, and Cloudflare R2 are the major examples.
Objects, keys, and buckets
You store an object (the data plus metadata) under a unique key within a bucket. Keys can contain slashes to simulate folders, but the structure is flat. Each object is retrieved by its key over a simple HTTP API.
How it differs from other storage
- Block storage backs disks and databases; it is low-level and fast.
- File storage offers a shared filesystem with directories.
- Object storage trades filesystem features for scale and durability.
Durability and tiers
Object stores replicate data for very high durability and offer storage classes from hot (frequent access) to cold archival. Lifecycle rules move objects to cheaper tiers as they age, optimizing cost automatically.
Common uses
Object storage is ideal for backups, data lakes, media, logs, and hosting static websites (often behind a CDN). It is not for low-latency random writes the way a database disk is; it is for durable, scalable blobs.
Role in CI/CD
Pipelines lean on object storage to store build artifacts and test reports, cache dependencies, and deploy static front ends by syncing a build folder to a bucket. A deploy step uploads assets and, if a CDN fronts the bucket, invalidates the cache so users see the new release.
Key takeaways
- Object storage keeps data as keyed objects in flat buckets, accessed over HTTP.
- It trades filesystem features for massive scale and very high durability.
- Pipelines use it for artifacts and to deploy static sites, often behind a CDN.