Formica Regains Control, Standardizes on VergeOS

A Brand That Became a Verb

Formica has made laminate since the company was founded in 1913. The brand name later became shorthand for the product category, the same way people reach for Google or Hoover. Today Formica operates worldwide as part of the Broadview Group, a Netherlands-based group of sister companies across the materials space.

Matt Collier is Formica’s Technology Architect, having spent 15 years with the company. He started on the help desk, moved into engineering, and stepped into architecture in 2019. Andy Hawes is the Service Delivery Manager for Europe and runs the infrastructure and help desk. The two carry a wide span of responsibility on a lean IT team. Matt framed the constraint directly.

“Formica cannot assign one person to the hypervisor and another to SQL. The team moves between systems and needs the environment to hold itself together.”

Reassessing the Platform

Formica ran a large on-premises VMware environment with StarWind vSan layered across the datacentre clusters. The design served the company well, and the team was not chasing change for its own sake. Broadcom’s acquisition of VMware prompted a review, and Formica used that moment to reassess what it wanted from its infrastructure. The bigger questions were operational. The team wanted less hand-holding, tighter integration between compute and storage, and a platform that could heal itself.

Under the former infrastructure, stability had its rough moments. Formica rebuilt its European StarWind cluster three times in one year after a design decision during an upgrade created a bottleneck that triggered timeouts across the infrastructure. That experience sharpened Formica’s interest in a single integrated platform with built-in resiliency, rather than a hypervisor and a separate storage layer stitched together.

Building the Short List

KVM was always the destination. Formica had ruled out Hyper-V due to management complexity, Nutanix on cost and hardware lock-in and set his sights on a KVM-based platform. The open question was which flavour and which vendor.

Formica evaluated a broad field. Nutanix came off the list early. Scale Computing demoed well and ran KVM underneath, yet it asked for reference hardware, and Matt wanted to move away from vendor-dictated hardware cycles. Red Hat OpenShift made the cut as another KVM option. XCP-ng fell out of favour due to the lack of enterprise support. Proxmox stayed in the picture for dev clusters, but its Austrian-time-zone support could not deliver the 24/7 follow-the-sun coverage Formica needed.

Formica set three requirements for the platform:

• One integrated product delivering both the hypervisor and the vSAN, from a single vendor
• Automation and self-healing, backed by resiliency through syncing and repair servers
• Strong overall value, without per-feature SKUs or nickel-and-dime licensing

The race came down to VergeOS and Red Hat, and it was close. What set VergeIO apart was enterprise support and a willingness to work as a partner. The company engaged VergeIO on a five-year commercial plan rather than handing over a fixed price and walking away. The all-you-can-eat model, with no per-feature SKUs, reinforced the fit.

Why VergeOS Won

Two things stood out in the evaluation. The first was hardware freedom. VergeOS did not demand reference hardware, so Formica could repurpose hardware it already owned. Matt’s summary was blunt.

“It could run it on a toaster if it had an internet connection. That flexibility kept serviceable hardware in production rather than on the scrap pile.”

The second was the demo. “When Andy and I sat down and demoed Verge, and saw how easy it was to deploy, I don’t think I’m wrong in saying we both just went wow,” Collier said. “The performance of the product, even on not-very-high-spec hardware, was impressive.”

Deployment in Minutes

Deployment starts simple. An administrator sets the name and IP address, picks the detected disks and NICs, and goes. “You go from having nothing to a working environment in less than 10 minutes,” Matt said. Formica leaned on VergeIO support through the first half-dozen builds to lock down the steps that follow, including certificates, groups, identity providers, and the vSAN. The team now runs deployments on its own.

Speed shows up in the field. A remote site reaches a configurable state in three to four hours, and most of that time is double-checking the configuration. Strip out the thinking time, and the real work runs about 90 minutes, and recipes can further trim that time. A Docker recipe delivers a VM with Docker already installed. LAMP stacks arrive ready to run. Each one saves time on every deployment.

Scaling to the Edge

Most of Formica’s footprint sits at the edge, well beyond the two data centres. Sites run across Europe and Asia, with North America now coming online, and each connects back to the core. Formica treats these edge locations as non-critical by design. A problem at one site stays contained to that site and does not take the business down. For a lean team supporting a global operation, that containment matters as much as raw uptime. It keeps a local failure from turning into a company-wide event.

Faster VMs, Less Hand-Holding

Performance moved in the right direction. Scheduled processing jobs that ran 40 to 50 minutes on VMware dropped to about 25 minutes on VergeOS. “Verge is very snappy and responsive,” Matt said. “Above all else, it seems to require less hand-holding than VMware.”

The integrated vSAN drove part of the improvement. VergeOS builds storage into the kernel rather than bolting it on as a separate product. Formica measured a speed advantage over StarWind with lower encryption overhead, and the work no longer ran on top of someone else’s hypervisor. He was impressed with VergeOS’s redundancy and resiliency.

Daily operational overhead fell off. “I can go days without having to look at the Verge cluster,” Matt said. Formica recently deployed Prometheus and Grafana for monitoring and added the VergeOS exporter plugin built by VergeIO. Alerts and dashboards now run on a platform that largely manages itself.

Andy pointed to the vSAN and the update process as the features he values most. Storage starts on a single node and grows without drama. Version updates run hands-off. VergeOS migrates workloads, places the first node in maintenance mode, reboots it, and repeats the sequence on the second node. “Press a button, download, apply update, hands off the steering wheel,” he said.

Matt put the overall verdict plainly.

“As a sysadmin, it needs to be easy and painless, and by and large it is. You’ve really got that right with the system.”

Resiliency and Backup by Design

Formica built resiliency into the architecture. Each of the two VergeOS data centres runs a production cluster and a repair server (ioGuardian). A VergeOS repair server is a synchronized remote system that holds a recent replica of a site’s data and steps in when a failure exceeds local redundancy, such as multiple drive failures across nodes. It automatically pulls back the missing or corrupted blocks from the remote copy, avoiding a snapshot rollback and keeping the system online.

The two data centres act as repair servers for each other, syncing back and forth so every copy lands offsite. Snapshot-based protection uses the built-in VergeOS backup and syncs to the second site.

The native tools eliminated the need for an additional management layer. Formica assessed other backup solutions after the move and found nothing that the built-in capabilities of VergeOS did not already cover. Adding a separate product would have meant one more layer to manage and additional costs for no real gain. Formica is evaluating a managed Verge deployment with an external vendor for longer term offsite retention of its snapshots to further improve its recovery posture.

Hardware That Lasts

Cost control extends into the hardware plan. In Taiwan, Formica procured new AMD EPYC clusters built to Verge-approved reference designs and validated the work with support. At the European data centre, the team lacked the budget for new hardware, so it followed the path VergeIO offered. It repurposed the existing hardware, added components, and extended its lifespan to 15 years. Those nodes run well and will stay in service until 2034.

Support That Delivers

Support earned steady marks. “The guys know their stuff,” Matt said. He has called outside normal business hours once and received the answer he needed without friction. Technical depth counts for more at Formica than polish, and VergeIO’s team brought it.

Looking Ahead

Formica continues its rollout. 13 sites run VergeOS today, including two data centre clusters. Three more edge sites are queued for North America, one of them in Canada, and the Ohio data centre will migrate off VMware and StarWind onto Verge. The team targets the end of 2027 to finish the remaining sites in Asia and Europe.

The Payoff

Formica set out to replace a platform, and it gained an operating model. One integrated system now carries compute and storage across every site, and it holds itself together with far less daily attention. The lean IT team spends its hours on projects instead of babysitting a hypervisor. Older hardware keeps earning its place, new sites come online in minutes, and the built-in resiliency protects data across its two data centres without a bolted-on backup product. For a global manufacturer running on a small team, VergeOS turned a difficult VMware decision into a durable operational advantage, and the same platform now carries the rollout through the rest of Asia, Europe, and North America.

“Everything’s great and we’re really happy with the solution.”

Matt Collier, Technology Architect, Formica