Block Storage

Block storage presents raw disk volumes to a server, which formats and mounts them like a physical hard drive — Amazon EBS and Google Persistent Disk are typical examples. It's distinct from object storage (S3-style), which serves whole files over HTTP. Why it matters: block storage is what you attach to a virtual machine to hold a database's data files, or anything needing low-latency random reads and writes and a real filesystem. It's fast and consistent, but tied to one instance at a time and to a single availability zone. Practical note: block volumes don't automatically survive the instance they're attached to — snapshot them regularly for backups, and remember that unused, unattached volumes keep billing quietly (a common FinOps cleanup target). Choose block storage for databases and stateful workloads; choose object storage for user uploads, backups, and static assets you serve to the world. Sizing and IOPS tiers directly affect both performance and cost.