Cloud Storage

Cloud providers offer several distinct storage types, each optimised for a different access pattern — from unstructured files to high-performance block storage attached to virtual machines.

Object, Block, and File Storage

Choosing the right storage type depends entirely on how the data will be accessed and by what kind of application.

• Object storage — stores files ("objects") with metadata, accessed over HTTP; ideal for unstructured data like images, videos, and backups (e.g. Amazon S3).
• Block storage — raw storage volumes attached directly to a virtual machine, like a traditional hard drive; ideal for databases and applications needing low-latency disk access.
• File storage — a shared file system accessible by multiple servers simultaneously, similar to a traditional network drive.

Storage Tiers and Redundancy

Most providers offer multiple storage tiers (e.g. "hot", "cool", "archive") trading off cost against retrieval speed, and typically replicate data across multiple physical locations automatically for durability.

Common Mistakes

• Using block storage for large amounts of unstructured file data, when object storage would be far more cost-effective and appropriate.
• Storing frequently accessed "hot" data in a cheap archive storage tier, resulting in slow retrieval times and unexpected retrieval costs.
• Assuming cloud storage is automatically backed up in a way that protects against accidental deletion — versioning and backup policies usually need to be configured explicitly.
• Choosing a storage type based on familiarity from on-premises systems, rather than the actual access pattern the application needs.