Block Storage
Definition
Storage that divides data into fixed-size blocks, similar to traditional hard drives, providing high performance for databases and applications.
Use Cases
- Netflix: Running stateful components and databases that require durable, low-latency disk attached to compute instances — Uses AWS infrastructure where block storage (EBS) is commonly used to provide persistent volumes for EC2-based workloads; data protection is typically handled with snapshots and replication patterns appropriate to the service (Enables reliable persistence and predictable performance for stateful services while keeping compute instances replaceable and scalable)
- Airbnb: Supporting production databases and other stateful workloads on cloud virtual machines — Uses AWS and commonly relies on EBS-backed storage for EC2 instances running stateful services; operational practices typically include monitoring disk performance, right-sizing volumes, and using snapshots for backup/restore workflows (Provides durable storage independent of instance lifecycle, improving operational resilience and simplifying recovery processes)
Provider Equivalents
- AWS: Amazon Elastic Block Store (EBS)
- Azure: Azure Managed Disks
- GCP: Google Persistent Disk
- OCI: OCI Block Volume
Frequently Asked Questions
- What's the difference between block storage and object storage?
- Block storage looks like a raw disk you attach to a server (you format it with a file system and read/write blocks). Object storage stores data as objects in buckets (accessed via an API/HTTP) and is better for files, backups, media, and large-scale unstructured data. Block storage is typically lower-latency for databases and VM disks; object storage is typically cheaper and scales massively.
- When should I use block storage?
- Use block storage when you need a disk-like device with consistent low latency and you control the file system or database layout. Common cases: VM boot volumes, database data files (SQL/NoSQL), transactional applications, container persistent volumes, and workloads needing high IOPS or predictable throughput.
- How much does block storage cost?
- Pricing is usually based on (1) provisioned capacity (GB per month), (2) performance tier (e.g., SSD vs HDD, provisioned IOPS/throughput), and (3) snapshots/backups stored. You may also pay for cross-zone/region replication or data transfer in some designs. Costs rise with higher IOPS/throughput tiers and frequent snapshots.
Category: data
Difficulty: intermediate
See Also