Concept
What Is Blob & Object Storage?
A blob (Binary Large OBject) is any opaque chunk of bytes — an image, video, PDF, ML model checkpoint, log archive, or database backup. Object storage is the system designed to hold billions of these blobs cheaply and durably, exposing them over a simple HTTP API. Amazon S3, Google Cloud Storage, and Azure Blob Storage are the canonical examples.
The core idea: instead of mounting a disk and navigating directories, you PUT a blob under a unique key and later GET it back by that key. There is no filesystem to manage, no capacity to provision, and no servers to patch — the storage scales effectively without bound and you pay only for what you store and transfer.
Why interviewers love it
- Almost every large system stores user uploads, media, or backups — so object storage shows up in nearly every design.
- It forces a clean discussion of the metadata-vs-blob split: store the bytes in object storage, store the searchable metadata (owner, key, size, content-type) in a database.
- It opens the door to high-value follow-ups: presigned URLs, CDN serving, storage tiers, and durability math.