Amazon FSx for NetApp ONTAP (FSx for ONTAP) has introduced a new capability to decrease solid-state drive (SSD) capacity. This capability, available in second-generation FSx for ONTAP file systems, gives users a way to reduce provisioned SSD capacity in place without data migration, application downtime, or file system rebuilds.
In this article, we detail how the SSD capacity decreasing capability works, its benefits, and best practices for implementation.
Here’s what I’ll cover:
- A leap forward in dynamic storage management
- The challenge: Why decreasing storage capacity isn’t so simple
- The solution: In-place SSD capacity decreasing
- Use case examples for decreasing SSD capacity
- Use case 1: Electronic design automation (EDA)
- Use case 2: Large-scale data migrations
- Use case 3: Media & entertainment (M&E)
- How it works: The technology behind SSD decrease
- Quick start: How to decrease your SSD capacity
- Monitoring the SSD decreasing process
- Key considerations and best practices
- In conclusion, right-size at will
A leap forward in dynamic storage management
The new ability to decrease SSD capacity is a significant step forward in cloud storage. While compute and other cloud resources can scale up and down seamlessly, volumes in some storage services cannot be reduced in size once they’re provisioned. With this new FSx for ONTAP capability, your storage capacity can be right-sized post-deployment—within a fully managed, highly available, and production-ready solution.
The impact extends beyond simple capacity management. Organizations running data-intensive, project-based workloads can now optimize costs and improve resource efficiency by eliminating the need to overprovision for peak demand periods.
The challenge: Why decreasing storage capacity isn’t so simple
Historically, reducing SSD capacity has involved tradeoffs between complexity and risk:
Option 1: Provision a new, smaller file system and migrate data volumes manually. This introduces downtime, requires extensive planning, and carries migration risks.
Option 2: Delete the existing file system entirely and rebuild with less storage. This is rarely viable for production environments, which have critical requirements for availability and data integrity.
These management challenges have created pain points for organizations with cyclical workloads where, typically, high-performance SSD capacity remains provisioned and billed despite reduced utilization. That can lead to longer migrations, a higher TCO, and added administrative challenges.
The solution: In-place SSD capacity decreasing
With in-place SSD capacity decreasing, FSx for ONTAP customers can now reduce SSD capacity on demand directly from the AWS Management Console or via the FSx for ONTAP API, without impacting application or data availability.
This capability seamlessly relocates data between SSD aggregates behind the scenes using the volume management functionalities in ONTAP and redirects client access, on a per-volume basis. This method abstracts away complexity for you and maintains uninterrupted access throughout the process. Beyond this operational efficiency, the capability unlocks other key benefits for storage management:
- Cost optimization: Pay only for the high-performance storage you need by provisioning storage for peak usage scenarios and scaling down afterward to reduce ongoing costs.
- Increased agility: Adapt efficiently to evolving requirements across distinct project phases without being locked into fixed, overprovisioned storage footprints.
- Accelerated data migrations: Improve performance by temporarily provisioning high-performance SSD tiers during data ingestion, then scaling back to lower-cost capacity tiers once the migration is complete.
The decreasing SSD capability integrates seamlessly with existing FSx for ONTAP management workflows, requiring minimal training or process changes for operations teams already familiar with the service. This makes it possible for your IT teams to avoid painstaking migration projects and spend their time on initiatives with more strategic value.
Use case examples for decreasing SSD capacity
Let’s take a look at some example use cases for the new capability.
Use case 1: Electronic design automation (EDA)
Chip design projects follow predictable cycles where massive SSD capacity supports intensive simulation workloads during active development phases. However, once testing is completed and a project moves to maintenance mode, the high-performance storage can be underutilized until it’s needed again, all while continuing to incur charges.
Traditionally, design teams would find themselves choosing between overprovisioning for peak periods or accepting performance constraints during critical phases. FSx for ONTAP SSD capacity decreasing lets them scale back and reduce costs without compromising on performance during peak phases.
Use case 2: Large-scale data migrations
Enterprise organizations regularly face large-scale data migration projects that can involve transferring terabytes or petabytes of data within tight time windows to minimize business disruption.
In such migrations, substantial high-performance SSD capacity is needed for testing and tuning various configurations. However, this extra capacity isn’t needed forever. After the initial setup phase, the data is compressed, deduplicated, and tiered to the capacity pool, at which point the SSD capacity is right-sized and the system is optimized. At that point, the extra capacity no longer serves any purpose. Traditionally, having this extra capacity on hand would mean choosing between compromising the system’s optimization or paying on an ongoing basis for unused storage.
With the ability to decrease SSD capacity with FSx for ONTAP, migration teams can provision optimal high-performance storage for fast, reliable execution, then scale down to actual operational requirements with the infrequently used data tiered to the capacity pool. This reduces both migration risk and ongoing costs, while eliminating the administrative overhead of manual data movement.
Use case 3: Media & entertainment (M&E)
Film production and gaming projects require substantial amounts of high-performance storage for rendering, visual effects processing, and asset management during active production. After project completion and archival, expensive SSD capacity sits idle until teams move to new projects.
Traditionally, production studios would maintain oversized storage allocations to avoid the complexity of manual migration processes to move data to new, smaller file systems. With SSD capacity decreasing on FSx for ONTAP, M&E studios can dynamically scale storage to fit each phase of production—without the burden of manual migrations.
How it works: The technology behind decreasing SSD capacity
The ability to decrease SSD capacity is based on NetApp® ONTAP® volume moving technology, a core feature that has provided reliable data migration functionality for decades in NetApp storage systems. This proven foundation brings enterprise-grade reliability to the cloud capacity reduction process.
The FSx for ONTAP background process follows a carefully orchestrated sequence:
- Initiating the decrease: When a decrease operation begins, FSx for ONTAP provisions a new, smaller set of SSDs to accommodate the reduced capacity requirement.
Note: In ONTAP terminology, an aggregate refers to a collection of physical storage devices (SSDs in this case) that are grouped together to provide storage capacity to volumes. When decreasing SSD capacity, FSx for ONTAP creates a new, smaller aggregate and migrates the volume data from the original aggregate to the new one. - Migrating the data in-place: An in-place data migration starts, moving data from the existing SSDs to the new storage aggregate on a per-volume basis.
- Maintaining service availability: Client I/O continues uninterrupted, with traffic seamlessly redirected to the new storage as each volume's data completes its migration.
- Deprovisioning old storage: Once all data in the SSD tier transfers to the new aggregate, the original storage aggregate is detached and deprovisioned.
This approach maintains data availability throughout the process while transitioning to the new capacity allocation and efficiently handles tiered storage configurations by focusing migration efforts on SSD-resident data. Data already stored in capacity pool tier is not physically moved. Instead, ONTAP moves metadata related to the data in capacity pool (a logical move). As a result, tiering data to the capacity pool has no impact on the SSD decrease completion time.
Quick start: How to decrease your SSD capacity
Initiating an SSD decrease operation is straightforward and available through multiple interfaces:
- The AWS Management Console provides an accessible approach for most users:
- Navigate to your FSx file system from the Amazon FSx console.
- Select the Update SSD storage capacity and IOPS option.
- Select Decrease as action type, and then specify the desired decrease amount in percentage or in absolute value.
d. The interface validates your request against system prerequisites before initiating the operation.
2. With the AWS Command Line Interface (AWS CLI) and API interfaces, you can unlock programmatic control via the update-file-system and UpdateFileSystem functionalities, respectively. This allows you to integrate SSD decreasing operations into automated workflows or infrastructure-as-code deployments.
3. Coming soon: BlueXP™ workload factory for AWS. This integration will allow initiating SSD decreasing from the workload factory interface, applying policies or workflows to automate SSD decreasing as part of lifecycle management.
Each interface provides the same core functionality with a different user experience to suit various operational preferences.
Monitoring the SSD decreasing process
Because the SSD decreasing capability runs in the background, monitoring is critical for transparency and troubleshooting.
Progress tracking is available through the Updates tab in the FSx for ONTAP console or via the describe-file-systems CLI or DescribeFileSystems API response. All three interfaces display the status, and the progress percentage with the final target capacity value.
This information helps operations teams plan related activities and respond to any issues that might require attention. There are three common scenarios that might require your intervention:
- Request rejected: This typically occurs when prerequisites aren't met.
To fix, make sure you meet all the prerequisites highlighted in the Key considerations and best practices section that follows. - Operation paused: This indicates SSD utilization in the new tier exceeded 80% during the migration process. Resolution options include tiering data to the capacity pool or removing data from volumes that have completed migration. The process resumes automatically once utilization drops below the threshold.
- Space needed: If additional space becomes necessary during a paused SSD decrease operation, you can submit a request to increase SSD capacity. The increase operation receives priority, and the decreasing process continues after the expansion completes.
Key considerations and best practices
Successful SSD decreasing operations benefit from careful planning and understanding of the process requirements and timeline. For a smooth SSD decreasing operation, keep in mind the following:
- Check the prerequisites: Are you ready to decrease your SSD capacity?
The SSD decreasing operation succeeds when specific conditions are met. Use this checklist to verify readiness:
- System support: This feature is available for second-generation FSx for ONTAP file systems. First-generation systems require migration to access this capability.
- SSD size: The file systems being decreased need to have at least 1,126 GiB of SSD capacity provisioned per high availability (HA) pair.
- SSD utilization: Current SSD utilization should be below 80%, and the new size should result in utilization remaining under 80%. This threshold provides sufficient space for staging writes during the migration process.
- Minimum decrease: The new size should be at least 9% smaller than the current size. Smaller reductions might be rejected to prevent frequent, inefficient operations.
- System state: All volumes should be online and accessible. Active clone relationships should be broken or deleted before initiating the SSD decreasing operation.
- Understand the timeline: This is a marathon, not a sprint
The SSD decreasing process duration varies depending on the size of the dataset involved—some operations can take hours, and others might require several days. This variability of duration is because the operation is designed to maintain data availability throughout the migration so that your data is always accessible and up to date for end-user applications. The only note is that you might experience a brief interruption or delay in I/O—lasting up to 60 seconds per volume—as client access is redirected to the new set of disks.
Expert tip: Because of the time and resources involved, SSD decreasing should be treated as a strategic, infrequent operation—ideal for two to three times per year as part of broader capacity and cost management planning.
3. Plan for performance: Give your file system headroom
Most workloads experience minimal impact during SSD decreasing operations, but write-intensive workloads might see temporary performance reductions. This reflects the resource requirements of the background migration process.
To best maintain consistent application performance during SSD decreasing operations:
- Plan to initiate the process during off-peak hours.
- Check that you have adequate file system headroom by keeping utilization below 50% for CPU, disk throughput, and SSD IOPS metrics before initiating SSD decrease operation.
- Consider temporarily increasing throughput capacity during the operation to offset the performance impact of the background migration process.
4. Know what you can’t do.
While an SSD decrease operation is in progress, you won’t be able to perform the following actions:
- Submit another request to decrease SSD capacity or SSD IOPS.
- Decrease provisioned throughput capacity.
- Add HA-pairs.
- Revert a volume to a previous state (volume snapshot restore) while data in the volume is being moved to the new set of disks.
- Run background storage efficiency tasks such as background deduplication and inactive data compression.
In conclusion, right-size at will
The new SSD decreasing capability for FSx ONTAP gives you greater control over high-performance file and block storage workloads on AWS - delivering lower costs and operational simplicity.
Whether you're managing dynamic EDA workloads, project-based media pipelines, or simply looking to improve your storage efficiency, this feature helps you move from overprovisioned to optimized—all on your schedule.
For detailed implementation steps and additional best practices for optimizing your SSD footprint, visit the Updating storage capacity and provisioned IOPS in the FSx for ONTAP documentation and start optimizing your SSD footprint today.
