NetApp has announced the general availability of its Ontap storage for VMware Cloud on AWS. According to NetApp, that will help customers make savings of 20-50% in TCO by being able to scale cloud storage independently of compute.
The offering is Amazon FSx for NetApp Ontap (FSxN) certified and supported as a supplemental datastore for VMware Cloud on AWS. NetApp said this is the first third-party storage provision for such workloads in AWS.
The move is a manifestation of NetApp and VMware’s ideas about multicloud operations, in which customers are able to run applications in their own datacentres, but also move enterprise workloads to potentially multiple clouds.
FSxN is NetApp Ontap storage in the AWS cloud, including SSD and via file access SMB and NFS protocols as well as block iSCSI.
FSxN for VMware Cloud was in private preview early in 2022, then public preview before this GA launch.
As for the other hyperscaler clouds, it is currently in public preview with Microsoft Azure and private preview with Google Cloud Platform.
Phil Brotherton, vice-president for solutions and alliances at NetApp, was asked when these would be generally available too.
“I can’t tell you,” he said. “That’s up to Microsoft and Google, as it was for Amazon and VMware to decide, because support starts with them.”
Brotherton explained how the offering would help customers achieve lower TCO for their VMware Cloud deployments.
“That’s because it is possible to scale the amount of storage,” he said.
“The source is the architecture. Previously, when you scaled storage and compute, it was at the same rate. When you separate compute and storage to scale independently, you can make savings on storage, but that varies depending on how high a storage load there is.”
But to what extent is NetApp seeing customers achieve multicloud working? While the idea might conjure up customer applications seamlessly operating between multiple on-site and cloud locations, most customers major on one cloud as a site to expand to, with others in use for testing.
To what extent could customers make use of something like FSxN for VMware across all three hyperscalers?
“It can work between the three,” said Brotherton. “But there are still frictions. Customers want the flexibility to choose which cloud, but most customers extend to one cloud from their datacentres. Which one they choose varies, and during prototyping, they will often test two at once. VMware reckons 70% of its customers use another cloud.”
What are the frictions Brotherton talks of when it comes to multicloud working?
“Networking and cost management, primarily,” he said. “You could set up work across different hyperscalers – and people do for availability reasons. But between hyperscalers, the cost efficiencies are difficult and it’s not economical, and that’s how most people make their decisions: Is it straightforward? Is it cost-effective?”
VMware has offered a cloud version of its vSphere virtualisation in the three big public clouds for several years. These are VMware cloud on AWS, Azure VMware Solutions and Google Cloud VMware Engine. Where that has included VMware vSAN storage, it has not been scalable independently.