Amazon EBS Volume Design and Operations - Selection Criteria for gp3 and io2 with Snapshot Strategies
Clarify the selection criteria for gp3 and io2 volume types, and learn practical IOPS/throughput design guidelines along with snapshot-based backup strategies.
EBS Volume Type Selection Criteria
EBS offers four categories: General Purpose SSD (gp3/gp2), Provisioned IOPS SSD (io2/io2 Block Express), Throughput Optimized HDD (st1), and Cold HDD (sc1). The starting point for selection is gp3. gp3 provides a baseline of 3,000 IOPS and 125 MiB/s at no additional cost, and can independently scale up to 16,000 IOPS and 1,000 MiB/s as needed. With gp2, IOPS was proportional to volume size, requiring larger volumes to increase IOPS, but gp3 removes this constraint. io2 Block Express is designed for database workloads and delivers up to 256,000 IOPS per volume. It is the only volume type offering a 99.999% annual durability SLA, making it suitable for mission-critical workloads like Oracle or SAP HANA. st1 is suited for big data and log processing workloads centered on sequential reads, providing up to 500 MiB/s of throughput at low cost.
IOPS and Throughput Design
Estimating IOPS requirements requires understanding the application's I/O patterns. Database-like workloads with heavy random I/O are IOPS-bound, while ETL and log analysis with sequential I/O are throughput-bound. Measure actual IOPS using CloudWatch's VolumeReadOps and VolumeWriteOps metrics; if VolumeQueueLength consistently exceeds 1, the volume is an I/O bottleneck. Adding IOPS to gp3 costs approximately $0.006 per IOPS/month (Tokyo region). Only consider migrating to io2 when requirements exceed 16,000 IOPS. io2 costs approximately $0.074 per IOPS/month, roughly 12 times more than gp3, so accurately estimating the required IOPS is the key to cost optimization.
Snapshot Strategy and Backup Design
EBS snapshots are incremental, storing only the blocks that changed since the previous snapshot in S3. The first snapshot copies all data, but subsequent snapshots capture only the differences, significantly reducing storage costs and creation time. Use Amazon Data Lifecycle Manager (DLM) to automate snapshots. DLM policies define tag-based schedules to automate daily and weekly snapshot creation and generation management. A common production configuration retains 7 daily snapshots and 4 weekly snapshots. Enabling cross-region copy provides DR coverage against regional failures. Fast Snapshot Restore (FSR) eliminates the latency penalty that occurs on first access to a volume restored from a snapshot. FSR is recommended for workloads like databases where first-access latency is unacceptable. For a systematic understanding of EBS from basics to advanced topics, books (Amazon) offer comprehensive coverage.
EBS Pricing
gp3 costs approximately $0.08 per GB/month, with 3,000 IOPS and 125 MB/s throughput included at no extra charge. It is 20% cheaper than gp2 (approximately $0.10/GB) and offers the advantage of independently configuring IOPS and throughput. io2 costs approximately $0.125 per GB/month plus approximately $0.065 per provisioned IOPS. st1 (Throughput Optimized HDD) is approximately $0.045/GB, and sc1 (Cold HDD) is approximately $0.015/GB. Use gp3 as the default choice and select io2 only when high IOPS is required to optimize costs.
Summary - EBS Design Best Practices
The fundamental approach to EBS volume design is to start with gp3, measure with CloudWatch metrics, and scale up as needed. Leverage gp3's flexibility to independently adjust IOPS and throughput, avoiding over-provisioning. Automate snapshots with DLM and prepare for DR with cross-region copies. Since volume type changes can be performed online, there is no need to aim for the optimal solution from the start. An iterative optimization approach based on measured data is the most effective strategy.