To understand Elastic Block Storage, we need to first understand what block storage is and how if differs from object storage.
Block storage breaks data into block and then stores those blocks as separate pieces. Each piece has a unique identifier. Object storage manages data as objects. Objects usually encapsulates relevant data. For example, a company is storing customer data in object storage. There are 50 data points collected about each customer and all are stored in the object. To update one data point, the entire 1Gb object needs to be modified and uploaded. If the data is stored in block storage, only the data point being modified needs to be changed.
Following are some features of block storage:
- You can choose SSD or HDD for your block storage
- Persistent (hard disk) and customizable block storage for EC2 instances (Elastic Cloud Compute)
- Block storage is replicated in the same Availability Zone (AZ)
- Option to encrypt blocks using AES-256
- Elastic volume
- Backup using snapshot
- Data can be backed up in point-in-time snapshots. EBS volumes can be restored from a snapshot.
- EBS is independent of EC2 instances
- You can detach an EBS volume from one EC2 and attach to another. If you lose your EC2 for whatever reason, your data remains safe in EBS.
- You can choose either RAID or LVM
- You pay for provisioned storage
- If you procure 1Tb but only use 100Gb, you will pay for 1Tb.
- Block are automatically replicated with the AZ and blocks are encrypted in transit.
keeping data durable throughout independent of the EC2 instance, connected over the network
pay for provisioned storage: if you procure 1Tb, you will pay for 1 Tb even if you only use 100Gb
can be used in RAID or LVM.