Careeroza

EBS Volumes

Amazon EBS (Elastic Block Store) provides high-performance, persistent block-storage volumes designed for use with Amazon EC2 instances. Think of an EBS volume as a virtual hard drive in the cloud that you can mount to a running server instance.

Unlike instance store disks, which wipe clean the moment an EC2 server powers down, EBS volumes persist independently. If your instance is stopped or suffers a hardware fault, the underlying storage data remains safely preserved.


Core Operational Architecture

  • Network-Attached Storage: EBS volumes are not physically slotted inside the same host machine housing your EC2 instance. They exist on a dedicated, distributed storage fabric and connect to your virtual machine over a highly optimized, low-latency private network link.

  • Availability Zone Constraint: A standard EBS volume is physically bound to a single specific Availability Zone (AZ). To mount an EBS volume to an EC2 instance, both resources must reside inside the exact same AZ. If you need to move data across zone boundaries, you must capture a snapshot and restore it as a new volume in the target zone.

  • Volume Attachment Ratios: By default, the mapping architecture follows a standard one-to-one or many-to-one relationship. A single EC2 instance can have multiple separate EBS volumes attached simultaneously (e.g., one drive for the OS root, one for database files, and one for media caches). However, a standard volume can only be mounted to a single server at a time.


Solid-State Drive (SSD) Volume Families

AWS splits SSD architectures into two primary branches depending on whether you need cost-effective, adaptable baselines or guaranteed high-throughput scaling limits:

General Purpose SSD (gp3) The baseline storage standard for the vast majority of cloud workloads. It decouples performance metrics entirely from physical disk size.

  • Operational Profile: Every gp3 volume includes a baseline performance allowance of 3,000 IOPS (Input/Output Operations Per Second) and 125 MB/s of throughput completely free, regardless of whether the volume size is 10 GB or 10 TB. If your application demands sharper performance, you can provision up to 16,000 IOPS and 1,000 MB/s independently for an incremental fee without paying for unneeded storage gigabytes.

  • Target Workloads: Virtual desktops, software development environments, test benches, virtual boot drives, and medium-sized relational databases.


Provisioned IOPS SSD (io2 / io2 Block Express) Engineered explicitly for high-end, mission-critical production systems that demand absolute performance stability and sub-millisecond latencies.

  • Operational Profile: Designed to consistently deliver up to 256,000 IOPS and 4,000 MB/s of throughput per individual volume with a maximum storage size ceiling of 64 TiB. It features a strict 99.999% volume durability rating (a 0.001% annual failure rate), making it 100 times more resilient against data degradation than general-purpose types.

  • Target Workloads: Large-scale enterprise transactional databases (such as SAP HANA, Oracle, or Microsoft SQL Server clusters) and high-frequency trading engines.


Hard Disk Drive (HDD) Volume Families

HDD storage families optimize for large-block sequential read and write speeds rather than rapid, random I/O operations. They cannot be used as boot volumes.

Throughput Optimized HDD (st1) A low-cost option designed entirely to maximize throughput bandwidth for heavy data streams.

  • Operational Profile: Focuses on shifting data rapidly in megabytes per second rather than tracking IOPS operations. It scales up to 500 MB/s of maximum sequential processing performance per volume.

  • Target Workloads: Big data parallel architectures (such as Apache EMR clusters), large-scale data warehousing, log analytics processing systems, and continuous media streaming workflows.

Cold HDD (sc1) The lowest-cost block storage option available on the EBS platform.

  • Operational Profile: Sacrifices access speeds and low-latency metrics to optimize purely for storage economics. It maxes out at a throughput limit of 250 MB/s.

  • Target Workloads: Cold, infrequently accessed archival data, backup compliance stores, or large staging partitions where real-time application access speeds are not a priority.


Advanced Architectural Capabilities

Elastic Volumes EBS features hot-swappable configuration boundaries. Without taking your EC2 instances offline or unmounting your filesystems, you can dynamically expand the capacity of a volume, transition between different volume families (e.g., converting a legacy gp2 drive to a modern gp3 layout), or scale up provisioned IOPS directly from the AWS Console or CLI. The changes take effect live while applications continue writing data blocks.

Snapshots and Archiving You can back up data by taking point-in-time EBS Snapshots, which are saved directly into Amazon S3. These snapshots are incremental—meaning only the specific blocks changed since your last snapshot are copied, keeping storage footprints light. If a backup file is kept purely for long-term compliance, you can move it to the EBS Snapshot Archive tier to drop your backup infrastructure costs by up to 75%.

EBS Multi-Attach Available on premium io2 families, this capability allows you to concurrently attach a single, shared EBS volume to up to 16 distinct Nitro-based EC2 instances within the exact same Availability Zone. Each instance maintains full read and write privileges to the shared block layer, which is essential for clustering corporate database engines that manage complex, coordinate-level disk fencing.