In a modern steel mill, operational continuity depends not only on furnaces and coilers but also on a digital layer that cannot fail. Big River Steel, one of the most advanced steel producers in the United States, operates with a constant demand: 24/7 production and systems capable of responding accurately and reliably. In this context, virtualization and storage infrastructure cease to be “corporate IT” and become a critical infrastructure, as crucial as any plant equipment.
At its facility in Osceola (Arkansas), with more than 1,200 employees, the company had relied for years on a widely adopted enterprise stack: VMware vSphere Enterprise Plus with vSAN, and later Dell VxRail for managing the platform lifecycle. This approach supported initial growth but gradually became a source of friction: more complexity, higher costs, and less flexibility.
The catalyst: rising costs and a sense of dependence
According to their own account, the turning point came when the environment started to feel overly dependent on third-party decisions: rigid licenses, reliance on proprietary components, and increased uncertainty following VMware’s transition under Broadcom (see article). In an industrial setting, this uncertainty translates into a concrete risk: planning risk.
This isn’t just about the cost to renew licenses. It’s about whether a critical platform can maintain predictability in budget, support, updates, and roadmap. When such predictability erodes, the debate shifts to a more philosophical idea: vendor independence.
The choice: Proxmox VE and a Ceph-based architecture
Big River Steel evaluated options and ultimately chose Proxmox Virtual Environment (Proxmox VE), driven by two factors that in industry weigh more than slogans:
- Long-term transparency and control, thanks to an open-source foundation and a more straightforward operating model.
- Integration with Ceph, to rebuild storage with a distributed and scalable architecture.
The migration was executed on a significant scale: 4 clusters of 3 nodes each (a total of 12 nodes) transferred from VMware to Proxmox VE. The team highlights that the process was carefully planned and trained, and that the transition was smooth without disruption.
What was before and what changed afterwards
Before the migration, Big River Steel operated:
- 2 clusters of 3 nodes based on Dell PowerEdge R740XD.
- 2 additional clusters of 3 nodes based on Dell VxRail P670F.
- An architecture utilizing vSAN, combining SSD, NVMe, and spinning disks, interconnected through multiple 25 GbE links.
After switching to Proxmox VE, storage was redesigned around Ceph, with organized pools (piscinas) differentiated—dedicated pools for SSD, NVMe, and standard disks—and CephFS for managing ISO images and templates.
The team’s evaluation is clear and endorsed by Brian Clark, Meltshop/PD Automation Engineer III: “We are extremely satisfied with Proxmox Virtual Environment. The seamless integration with Ceph was a real turning point, giving us the flexibility and scalability we needed without the high licensing costs of vSAN.”
Comparison table: “before and after” in the Big River Steel case
| Area | Before (VMware) | After (Proxmox VE) |
|---|---|---|
| Virtualization | vSphere Enterprise Plus | Proxmox VE |
| Storage | vSAN (HCI) | Ceph (with SSD/NVMe/standard pools) + CephFS |
| Hardware referenced | R740XD + VxRail P670F | Reuse of existing environment (same hardware) |
| Networking referenced | Multiple 25 GbE links | Maintains high-capacity networking approach |
| Operational goal | Growth with increased dependence and complexity | More predictable, flexible, and resilient platform |
Reduced firefighting: impact on daily operations
In industrial virtualization, success isn’t measured by a demo but by what happens during midnight shifts. This case illustrates a tangible change: greater stability and performance, leading directly to less time spent troubleshooting and more time dedicated to process improvement and innovation.
Additionally, Big River Steel implemented Proxmox Data Center Manager to gain a comprehensive view of their clusters and centralize management. The team emphasizes the practical usefulness of the interface and its integration into daily operations—a layer that becomes meaningful when infrastructure shifts from being “a cluster” to a set of domains to be governed consistently.
Implications of this case for the debate on vendor independence
The Big River Steel example helps ground a recurring IT debate: Is vendor independence a luxury or a necessity? In mission-critical environments, it often functions as an operational insurance policy offering three concrete benefits:
- Predictability: enhanced capacity to plan costs and technological evolution.
- Optionability: real ability to renegotiate, scale, or change without rebuilding the entire stack.
- Explicit architecture: clearer visibility of technical decisions (and thus more auditability and sustainability).
At the same time, it’s important to recognize the flip side: independence doesn’t mean “doing everything alone.” In this case, the company highlights the value of vendor support and official training to adopt best practices and gain operational confidence. Control is built both with technology and with internal capabilities.
Frequently asked questions
What types of companies should consider migrating from VMware to Proxmox VE?
Organizations with mature virtualized infrastructure—especially those seeking predictable licensing, technical flexibility, and lifecycle control—often evaluate alternatives. In 24/7 sectors, the key decision factor is usually operational continuity and planning capacity.
Can Ceph replace vSAN in hyperconverged architectures?
Ceph can serve as distributed storage in virtualized environments, with design options (performance pools, replication, file systems like CephFS) that adapt architecture to various workload profiles. Its suitability depends on design, network, operational practices, and team expertise.
What is “vendor lock-in” and why is it more concerning in industrial environments?
Vendor lock-in refers to technical and contractual dependence on a single provider to sustain critical systems. In industry, lock-in escalates risk when licensing, support, or roadmap terms change, because the cost of failure or delayed reaction is higher.
What role do training and support play in critical virtualization migrations?
They are central. In this case, the team emphasizes that support and training helped adopt best practices and operate the new environment with confidence, minimizing friction and reducing risks during the transition.
Source: Proxmox Solutions