3.2 High-Throughput Storage: S3 Partitioning, EBS gp3/io2 & EFS
Design high-performance storage solutions selecting between Amazon S3 partitioned prefixes, EBS gp3/io2 Block Express, and Amazon EFS shared file systems.
🎯 Key AWS SAA-C03 Architectural Takeaways
- Amazon S3 scales horizontally to support at least 3,500 PUT/POST/DELETE and 5,500 GET requests per second per partitioned prefix.
- S3 Transfer Acceleration speeds up global uploads by routing data over CloudFront edge locations, and S3 Multipart Upload enables parallel part uploading.
- Amazon EBS gp3 provides baseline 3,000 IOPS and 125 MB/s independent of volume size; io2 Block Express delivers up to 256,000 IOPS and 4,000 MB/s throughput with 99.999% durability.
- Amazon EFS provides elastic NFSv4 shared storage across multi-AZs; Max I/O mode supports massive concurrent read/write scaling for container clusters.
Storage architecture selection dictates application throughput and scalability. Amazon S3 automatically creates storage partitions based on object key prefixes. Each unique prefix supports 3,500 PUT/POST and 5,500 GET requests per second. Applications with extreme concurrency must distribute keys across multiple prefixes (using dates, hashes, or UUID prefixes) to achieve linear horizontal scaling.
For high-volume global uploads, combining S3 Transfer Acceleration with S3 Multipart Upload is best practice. Transfer Acceleration routes traffic onto AWS private fiber via edge locations, while Multipart Upload splits files larger than 100 MB into chunks that upload in parallel. If a transient network glitch drops a packet, only the affected part is retried rather than restarting a multi-gigabyte upload.
For block storage attached to Amazon EC2, EBS gp3 volumes offer 3,000 baseline IOPS and 125 MB/s throughput at any storage size, with the ability to provision up to 16,000 IOPS independently. For mission-critical transactional databases requiring extreme I/O, EBS io2 Block Express scales up to 256,000 IOPS and 4,000 MB/s throughput per volume with sub-millisecond latency and 99.999% durability.
When multiple compute instances or containers require concurrent POSIX file access, Amazon Elastic File System (EFS) provides a shared NFSv4 filesystem spanning multiple Availability Zones. EFS offers General Purpose performance mode for low-latency web serving and Max I/O mode for big data analysis, media processing, and massively parallel container workloads.
⚠️ Common Pearson VUE / AWS Exam Traps
- Placing all uploaded objects under a single S3 folder prefix at high request volumes, triggering HTTP 503 Slow Down partition throttling.
- Selecting EBS gp2 when independent IOPS and throughput provisioning is needed — gp3 decouples storage capacity from IOPS and throughput.
- Using standard EBS volumes when multi-instance shared file access is required across multiple Availability Zones — EBS is single-AZ; use EFS.
An enterprise is migrating an on-premises Oracle database to an Amazon EC2 instance. The database requires sustained storage performance of 80,000 IOPS and 2,000 MB/s throughput with consistent sub-millisecond I/O latency and 99.999% volume durability. Which Amazon EBS volume type fulfills these performance and reliability requirements?