VMware Inc. today is breaking out the technology it uses for its Cloud Disaster Recovery service and selling it as a cloud storage service.

Cloud Flex Storage is described as a scalable, elastic and cloud-native offering that is fully managed by VMware and delivered with pay-as-you-go pricing on the VMware Cloud on Amazon Web Services Inc. infrastructure. Customers can provision the storage quickly in the VMware Cloud Services Console without adding hosts and adjust their storage capacity as needed.

“This is about scalability and elasticity with scaling disaggregated so you don’t have to add more hosts to get more storage capacity,” said Mark Chuang, head of product marketing for cloud storage and data.

The elimination of the need to add hosts is important because VMware has always sold cloud storage on a hyperconverged infrastructure model that required customers to buy additional central processing unit capacity in order to get more storage space.

The service is built on a Log-Structure Filesystem that was developed by Datrium Inc., which VMware acquired in July 2020. Its two-tier design has both an immutable layer and a caching layer based on high-performance Non-Volatile Memory Express drives with an object storage back end built on AWS’ S3 storage service. This enables both speed and data protection and is optimized for uses such as backup, disaster recovery, ransomware protection and data recovery, VMware said.

Currently available as a preview edition for approved customers, VMware Cloud Flex Storage supports the lifting and shifting of virtual machines without the need to rework the data layer or re-architect the storage design, a time-consuming process that is typical of cloud migrations.

Customers that use both VMware’s vSAN on-premises storage services can use Cloud Flex to spread their data across all three storage platforms and manage it with the VMware vCenter console.

Expansion plans

The company said it plans to deliver the same storage option across multiple clouds and to enable multidimensional scaling of compute, performance and storage capacity as well as adding data management and data reduction capabilities across multiple regions.

“The vision is to concentrate on supporting a broad set of workloads with different performance characteristics as well as data types,” Chuang said. The initial offering is limited to network filesystem data and does not support such advanced data management functions as cloning and snapshotting.

This release also doesn’t support use by cloud-native containers, although that is on the roadmap, Chuang said.  Customers can provision up to 400 terabytes of usable capacity per data store, with data encrypted at rest.

“A key design intent is to make this a cost-effective but high-performance option,” Chuang said. “We believe it can be more cost-effective than vSAN storage because it can add storage capacity without needing to also add compute capacity.”

Asked if that meant Cloud Flex could eventually supplant vSAN, Chuang said there are plans to so that to meet different application requirements but VMware isn’t sharing any timing details at this time. “Today it is supplemental to vSAN but we do envision a scenario of being able to enable diskless hosts with Flex Storage as the primary storage offering,” he said.

VMware declined to say when the preview will end or additional cloud platforms supported. Pricing was also not announced.

Photo: Robert Hof/SiliconANGLE