Tech ONTAP Blogs

Amazon FSx for NetApp ONTAP doubles performance

MikeMcNamara
NetApp
3,585 Views

 

Over the past 2 years, Amazon FSx for NetApp ONTAP has consistently released new features to support a wide variety of workloads. You get support for workloads such as user and application file shares, relational databases (SAP HANA), datastores for VMware Cloud on AWS, and backup and disaster recovery. FSx for ONTAP offers fully managed, feature-rich NetApp® ONTAP® file systems in the cloud. You can easily migrate or extend your workloads that rely on the widely used data management features of ONTAP, such as NetApp Snapshot™, SnapMirror®, and FlexGroup technology. And you don’t have to modify your applications, tools, processes, or workflows.

 

At AWS re:Invent 2023, FSx for ONTAP scale-out file systems was introduced to support an even broader range of high-performance workloads on AWS. Scale-out file systems deliver the performance of multiple file systems in one by automatically spreading your workloads across multiple highly available (HA) pairs of file servers. This enabled you to easily lift-and-shift even more high-performance ONTAP workloads to AWS with up to 36 GB/s of throughput and 1.2M IOPS.

 

But customers with the most compute and data intensive ONTAP workloads—such as large-scale electronic design automation (EDA), AI/ML and generative AI (GenAI), seismic analysis, and visual effects (VFX)—may require even higher performance. If you’re one of those customers, the latest FSx for ONTAP enhancement has you covered.

 

FSx for ONTAP scale-out to 12 HA pairs

Responding to customer feedback, AWS and NetApp doubled the FSx for ONTAP capabilities. The service can now support up to 72GBps of read throughput, 12GBps of write throughput, and up to 2.4 million SSD IOPS with <1ms latency and a 1PB-scale namespace. Now you can spread your workloads across up to 12 pairs of file servers to support an even wider array of high-performance workloads in AWS. You can easily lift and shift your workloads to AWS without having to rearchitect to take advantage of the agility, scalability, and security of the cloud. With this enhancement, you can efficiently manage, process, and analyze large volumes of data across different AI/ML workflows and other compute-intensive workloads, while maintaining enterprise-class data management.

 

As the demand for these compute and data intensive workloads grows, the need for high-performance, low-latency data-processing capabilities also multiplies. Proven data management features, such as incremental Snapshot copies, clones, cross-region replication, and data-reduction techniques like compression and deduplication, along with fully elastic capacity, have all become essential. These capabilities help your organization improve performance and scalability, maintain data uptime and protection, reduce costs, and seamlessly migrate data to the cloud.

 

MikeMcNamara_0-1709567576583.png

 

FSx for ONTAP benefits

FSx for ONTAP scales to petabytes in size, enabling you to store virtually unlimited unstructured data in a single namespace. Features like NetApp Snapshot technology creates point-in-time copies of your data to protect against data loss. And with SnapMirror data replication, FSx for ONTAP can copy your data to a backup file system across regions and recover it if data loss or a disaster hits. Multi-AZ high availability protects against zonal failures and provides the lowest possible recovery objectives.

 

FSx for ONTAP supports multiprotocol access (NFS, SMB, iSCSI, S3) to the same data and is tightly integrated with AWS Identity and Access Management to maintain your data security, accessibility, and mobility. With NetApp FlexClone® technology, you can create instant, cost-free, point-in-time clones of data volumes. So you can copy even the largest datasets almost instantaneously, significantly accelerating your development and testing (DevTest) processes. FlexClone is particularly useful for facilitating rapid iteration for AI/ML and DevTest use cases. For example, for AI/ML, you might need multiple copies of identical datasets (such as for model training and testing or data augmentation). Or DevTest might require temporary copies of a dataset (for experimentation and testing an application against a production dataset).

 

Cost savings and hybrid cloud support are other notable FSx for ONTAP advantages. You get automatic data tiering, thin clones, and storage efficiency features like thin provisioning and data deduplication to cut your storage costs. And its interoperability with on-premises ONTAP file systems enables you to build hybrid environments by using your own foundational models in the cloud.

 

FSx for ONTAP customer examples

Let’s look at how a broad range of customers have harnessed the power of FSx for ONTAP to drive innovative solutions and to achieve strong outcomes. COVVI is a company that’s dedicated to empowering individuals who are missing upper limbs. Vaultedge is a document AI platform for the banking and insurance sectors. Arcesium provides advanced data capabilities for the financial services industry. And ARM is a leader in CPU technology.

 

COVVI provides people who have missing upper limbs with the ability to perform tasks like preparing a meal or holding hands with a loved one. The COVVI Hand marries speed, dexterity, and strength to naturally mimic a hand’s complex gestures, grips, and manipulations as closely as possible—and it’s all managed remotely through an app.

 

The initial use case for FSx for ONTAP at COVVI was for the company’s R&D files, which had previously lived unsecured on personal computers and workstations. COVVI has also added a data lake where incoming data from the users’ hands are sent back for analytics. From that data, AI and ML techniques are employed to learn the gestures and movements from everyday use of prosthetic hands. COVVI can then update the software on the prosthetic hand remotely, adding new movements and resolving software bugs in real time.  

 

Vaultedge is a document AI platform that’s built on a large language model architecture for the banking and insurance industries. Customers use Vaultedge to automate document-processing workflows as part of loan processing, income underwriting, insurance underwriting, insurance claims processing, and other related tasks. Vaultedge needed a shared file system that was highly available; that provided the necessary throughput and performance; and that performed backup and replication, but also with storage efficiency features like deduplication and compression.

 

After evaluating multiple offerings, Vaultedge chose FSx for ONTAP because it provided an enterprise-grade shared file system to process their customer data without having to worry about performance or data loss. The ONTAP data management and data efficiency features have lowered Vaultedge’s costs, and with NetApp SnapMirror technology, the company has a reliable replication and disaster recovery solution.

 

Arcesium designs advanced data, operations, and analytics capabilities for the financial services industry and is used by some of the world’s most sophisticated asset managers and financial services organizations. Arcesium needed to reduce the time and costs associated with refreshing mission-critical, multiterabyte databases and to maintain a consistent point-in-time view across multiple databases. With FSx for ONTAP, NetApp FlexClone® volumes enabled Arcesium to create near-instantaneous, space-efficient database copies for system refresh purposes. And as a managed service, FSx for ONTAP was easy to migrate to, was simple to operationalize, and reduced operational complexities.

 

A quote from Arm in a NetApp press release during AWS re:Invent 2023 highlights the positive experience that Arm has had with FSx for ONTAP. “Keeping pace with the rapid rate of cloud innovation requires agile, performant, and secure solutions,” said David Miller, senior director and fellow of IT Architecture at Arm. “In testing the new Amazon FSx for NetApp ONTAP, we’re successfully leveraging the benefits of the cloud to unlock a new set of scale-out storage capabilities and higher performance for our internal HPC workloads.”

 

FSx for ONTAP to unlock your organization’s full potential

FSx for ONTAP keeps pace with rapidly advancing workloads by providing high-performance storage that doesn’t sacrifice speed, agility, or efficiency. With FSx for ONTAP 12-HA-pair scale-out file systems, you can elevate your GenAI, EDA, VFX, seismic analysis, and other compute and data intensive workload storage capabilities to new heights, unlocking the full potential of these technologies. To learn more, read the FSx for ONTAP user guide and FSx for ONTAP tactical buyers’ guide.

Comments
Public