Cloud Volumes Service is NetApp’s new, fully managed hyperscale cloud service for enterprise users seeking high performance and reliability.

Cloud storage comes in many forms, from object storage to block-level devices. However, many organizations simply require a ready-to-use, highly available, high-performance shared file service where scalability is not an issue.

This is the solution that NetApp® Cloud Volumes Service offers as a native cloud marketplace service in AWS and Google Cloud Platform, and as a first-party service called Azure NetApp Files, delivered supported by Microsoft, built on NetApp technology in Microsoft Azure.

By leveraging artificial intelligence (AI) in their applications, organizations can quickly make data-driven decisions to drive efficiencies in their products and services. The cloud is an indispensable platform that not only allows one to quickly experiment with ideas via proof-ofconcepts, but also provides a rich software ecosystem for building AI applications and training deep learning models.

In addition to compute engines optimized for massively parallel computations on the cloud, you need a high performing file system to work seamlessly with GPUs and keep them busy at ~100% utilization. 

Artificial intelligence (AI) is increasingly being used in multiple industry verticals to support important business needs such as automating business processes, providing cognitive insights through data analysis, and interacting with customers using natural language processing. Deep learning (DL), a version of machine learning, is effective at learning from large volumes of data to mimic the human brain’s pattern recognition (for example, images, speech, and text).

Organizations face many challenges in using AI in their cognitive projects, and one of the top challenges is choosing the right set of technologies and infrastructure. The technologies and infrastructure influence the models’ prediction accuracy, which is a function of the dataset size used to train them. And speed is an issue too: depending on the complexity of the neural network and the volume of datasets, model training can run from a few hours to days.

Cloud Volumes Service

NetApp Cloud Volumes Service is a fully managed, cloud-native file storage service based on the proven NetApp ONTAP® data management software and other data services. Cloud Volumes Service combines NetApp’s vast file services expertise with the simplicity and flexibility of the biggest clouds (AWS, Azure, Google). It also makes it easier to integrate with the ecosystem of applications in the cloud. The service supports NFSv3 and NFSv4 (coming soon) for Linux/UNIX clients, and SMB for Windows clients operating in the cloud.

Issues with Existing NFS/Shared File Solutions on the Cloud

The advent of hybrid cloud architectures and introduction of containers has simplified application and workload migration between public clouds and on-premises environments. However, the lack of robust file storage services offered by cloud providers inhibits data mobility.

  • Nearly 80% of all file data is created through NFSv3, yet there are few options in the cloud that match the file offerings available in on-premises file storage systems. 
  • Applications storing file data demand high file storage performance, which can often be an expensive proposition in the cloud. 
  • Issues around achieving data consistency between on-premises environments and the public cloud can result in costs for data ingress and egress.

Benefits of Cloud Volumes Service

Cloud Volumes Service delivers fast file-storage performance and business-critical file services to cloud applications. This service harnesses NetApp’s file services expertise and addresses key issues with cloud-based file storage services. Here are a few key value propositions of this service:

  • Cloud Volumes Service is offered in all the top public clouds that support NFS and SMB.
  • It is delivered as a simple native cloud service, giving you an easy-to-use interface and a single payment model through the cloud provider. 
  • It provides features to migrate, replicate, and synchronize data across on-premises environments and the cloud. 
  • You can use NetApp SnapshotTM copies for data protection and data restores. 
  • Built-in, always-on encryption allows protection without a performance impact. 
  • You can scale from zero to 100TB deployments in a matter of seconds.

Cloud Volumes Service on AWS

NetApp Cloud Volumes Service for AWS is designed to increase performance while reducing complexity. The multiprotocol storage service natively running in AWS simplifies data migrations for on-premises environments. It supports NFSv3, NFSv4 (coming soon), and SMB. Being a fully managed service, it handles configuring and managing all storage infrastructure. This service is a great fit for developers, lineof-business (LOB) engineers, database administrators, and application architects who consume storage capacity but do not want to administer it. Soon, Cloud Volumes Service will also be integrated with other AWS services, such as Amazon Elastic MapReduce (EMR), Amazon EC2, and Amazon Elastic Container Service (ECS).

The cloud provides a rich software ecosystem to develop, deploy, and test AI/DL applications. It is common for small to large organizations to get started on the cloud by using GPUs for model training. GPUs are priced resources and need to be maintained at high utilizations in order to reduce training times, allow for faster experimentation, and as a result, minimize the cost of usage. A high-performance, easy to use file system that will prevent GPUs from waiting for data is imperative in accelerating model training on the cloud and to optimize cost.

NetApp Cloud Volumes Service fits perfectly for DL model training on the cloud. It is a fully managed, cloud-native file storage service with an intuitive user interface (and CLI) that will enable you to get started quickly and deliver the high performance required by data intensive AI / DL applications. The Cloud Volumes Service, uniquely optimized to accelerate model training on the cloud, thus reducing costs is ideally suited for organizations starting their AI journey on the cloud as well as the ones with existing cloud deployments for AI.

Download Whitepaper for learn how to create volumes for NFS and SMB clients, create snapshots, use Cloud Volume APIs, and clone volumes and replicate data with Cloud Volumes Service for AWS.

To read full download the whitepaper:
AI Deployment on the Cloud with NetApp Cloud Volumes Service