sborisov - Fotolia

IaaS firm uses Pure Storage as a service for ‘virtual datacentres’

Espresso Gridpoint moved from largely NetApp to Pure Storage FlashBlade file and object for its ‘virtual datacentres’ business model that offers VMware, Hyper-V and ProxMox VMs

Dutch infrastructure as a service (IaaS) provider Espresso Gridpoint has shifted to Pure Storage Evergreen//One storage as a service to support its virtual datacentre business model, and reaped the benefits of lower costs, proactive support and upgrades, as well as reduced energy costs.

Gouda-based Espresso Gridpoint had run some Pure Storage hardware since 2019, but gradually shifted to major on Pure for storage hardware, where it previously depended heavily on NetApp storage.

Espresso Gridpoint delivers “virtual datacentres” to its customers – MSPs and software suppliers – who sell those services on to their clients. End customers span sectors that include legal, education, local government, healthcare, accountancy and transport.

Its hardware comprises HPE compute, Cumulus and Nokia networking equipment, with Pure Storage FlashBlade file and object storage. From this infrastructure it delivers virtual machines from Hyper-V, VMware and ProxMox, with a lot of movement occurring between the latter two recently, according to CEO Marga Reuver.

According to Reuver, the company’s storage volumes now held in Pure Storage are measured “in the petabytes”, with a move planned from FlashBlade//E to FlashBlade//S arrays in January 2025.

FlashBlade is Pure’s fast file and object storage family, which provides rapid access to large-capacity storage. It’s aimed at what have traditionally been secondary storage use cases – backup, archive, analytics datastores – but which have evolved to require rapid input/output (I/O) for analysis, recovery and so on.

FlashBlade//S can provide capacity and high performance in a range of models. FlashBlade//E provides capacity-focused QLC media aimed at unstructured data use cases where performance is less important.

Read more about consumption models of storage purchasing

Espresso Gridpoint moved to Pure’s Evergreen//One storage-as-a-service model last year, said Reuver.

Consumption models of IT and storage procurement have become increasingly popular. Since the emergence of the cloud, with pay-as-you-go as a norm, customers have come to expect flexibility in use, deployment, upgrades, scalability, speed of development and roll-out in their infrastructure.

And so, storage suppliers have adapted. Procurement options now range from full ownership with lifetime upgrades to pay-as-you-go with storage capacity and performance upgrades triggered via AIOps monitoring.

“In our world, capex vs opex is important, and Pure Storage is able to fully support its hardware in a service programme,” said Reuver. “Also with other vendors, such as HPE, it’s more of a buy-and-lease arrangement. With Pure, you get everything, including upgrades, and moves to different configurations when we want to expand.

“The ability to scale up and down in storage capacity is important for us,” she added. “As an infrastructure provider it is vital to be able to predict that kind of thing and respond to it.”

Reduced cost

According to Reuver, key benefits come via reduced power consumption, better and responsive support from the supplier, but also reduced cost per GB of storage, which it can pass to the customer.

“It means we can offer services to our customers at prices that come in below those of the public cloud providers,” she said. “And that makes it cheaper for their customers.

“We need to make money and achieve our margins,” added Reuver. “And it’s definitely cheaper than NetApp. It’s scalable, reliable and can handle huge volumes of data.”

One customer that uses Espresso Gridpoint’s services is Groningen-based AI-InfraSolutions, which uses mobile mapping cars to capture petabytes of geospatial data annually. It handles up to 12TB of data a day via its data lake and, according to Reuver, was able to cut processing times for that volume of information from 38 hours to 22 when Pure Storage was deployed.

Read more on Datacentre capacity planning