Amazon EBS
A block storage service that attaches to EC2 instances, offering multiple volume types including SSD and HDD to match different workload requirements
Overview
Amazon Elastic Block Store (EBS) is a high-performance block storage service used with EC2 instances. It serves as storage for file systems and databases, and data persists even when instances are stopped or restarted. The four main volume types are General Purpose SSD (gp3), Provisioned IOPS SSD (io2), Throughput Optimized HDD (st1), and Cold HDD (sc1), allowing you to choose based on the balance of IOPS performance, throughput, and cost. EBS volumes are automatically replicated within the same Availability Zone, protecting data from single hardware failures. The snapshot feature lets you save point-in-time backups of volumes to S3 and restore them in a different Availability Zone or region.
gp3 and io2 - Volume Selection Based on IOPS Requirements
EBS volume selection works backward from the IOPS and throughput numbers your workload requires. gp3 (General Purpose SSD) includes baseline IOPS and throughput in its price, and can be scaled beyond that baseline for an additional fee. Because the IOPS ceiling scales with capacity, however, allocating IOPS near the top of the range requires a correspondingly large volume. Baseline and ceiling values are defined per volume type, so check the AWS documentation for the values that apply when you design. Web server root volumes, small to mid-size RDS instances, and most development environments are well served by gp3. Migrating from gp2 to gp3 can be done online and can lower costs while maintaining equivalent or better performance. io2 Block Express, on the other hand, offers a higher IOPS ceiling than gp3 and the sub-millisecond latency required by large-scale databases such as Oracle and SAP HANA. Azure Managed Disks also supports IOPS settings independent of capacity through Premium SSD v2 and Ultra Disk, and the difference between the two platforms shows up in operating conditions rather than in ceiling values. Ultra Disk and Premium SSD v2 cannot be used as OS disks and are data-disk only, whereas io2 can also be selected for an EC2 root volume. The billing models differ as well: io2 charges for capacity and provisioned IOPS on a usage basis, while Ultra Disk provisions capacity, IOPS, and throughput separately and additionally requires Ultra Disk support to be enabled on the VM. Ultra Disk is also constrained by which regions, VM sizes, and availability zones are supported in combination, so when evaluating a migration you should check these conditions rather than only the headline specs. Among the HDD types, st1 suits sequential reads for log processing and big data, while sc1 fits infrequently accessed archive data and keeps the per-GB price substantially lower.
Snapshot and Encryption Design
EBS snapshots use an incremental backup approach that stores only changed blocks, minimizing both storage costs and creation time. You can restore snapshots in a different Availability Zone or region, making them a core building block for disaster recovery strategies. Scheduling regular snapshots through Amazon Data Lifecycle Manager automates backup retention and cleanup without manual intervention. EBS encryption should be enabled by default, using KMS customer-managed keys so that key rotation and access control can be managed according to organizational policies. Enabling default encryption at the account level ensures that every new volume is automatically encrypted, eliminating the risk of unencrypted volumes being created by oversight.
Cost Optimization and Sizing Considerations
EBS cost optimization starts with right-sizing volumes. Migrating from gp2 to gp3 can lower costs while maintaining equivalent or better performance, since gp3 decouples IOPS and throughput from volume size. Monitoring actual IOPS and throughput utilization through CloudWatch metrics helps identify over-provisioned volumes that can be downsized. For io2 volumes, provisioning only the IOPS actually needed rather than peak theoretical demand significantly reduces monthly costs. Transitioning infrequently accessed snapshots to the EBS Snapshots Archive tier lowers snapshot storage costs. When designing volume sizes, account for growth headroom but avoid excessive over-provisioning - EBS Elastic Volumes lets you resize, change type, and adjust performance of live volumes without downtime, so you can start conservatively and scale up as needed.
References (Official AWS Resources)
The primary sources for this page are the official AWS website and documentation. Check the official pages below for the latest specifications and pricing.
If this page and the official documentation disagree, treat the official documentation as authoritative.