CIO / VP of IT
A predictable infrastructure foundation that grows without repeated hardware refreshes or a headcount curve to match.
Service · Infrastructure as a Service (IaaS)
Managed private cloud on Proxmox VE and Ceph.
You keep your operating systems, applications and data. ModernOps runs the clustered infrastructure underneath them, including monitoring, capacity and hardware refresh, on a production private cloud operated from our own facilities.
Infrastructure / Virtualization Leader
Enterprise clustering, HA, tenant isolation, storage and lifecycle operations in one operating model.
IT Operations Lead
Provisioning, incidents, changes and escalation through documented workflows and scheduled service reviews.
Finance / Strategy Leader
Explicit consumption units, matched to measured workload demand, as an alternative to per-core hypervisor licensing.
The decision
Key design
considerations.
Five questions come up in most hosted-platform conversations, and the answers decide the architecture.
01 /
Hypervisor renewals changed the decision.
Per-core hypervisor licensing turned the hardware refresh into a platform decision. A managed private cloud is one exit that keeps the clustering and HA capabilities, removes the capital cycle, and bills in units your finance team can read, without moving to a hyperscaler.
02 /
Where the responsibility line sits.
You control operating systems, applications and data. We operate the infrastructure, availability, capacity and platform health underneath them. The responsibility table further down this page spells the line out.
03 /
Public cloud and private cloud have different jobs.
Public cloud suits elastic and exploratory demand. For steady-state virtual machines, a private platform billed in flat units (vCPU, memory in GiB, storage in GiB) produces an invoice that changes only when the workload does.
04 /
Shared platform, isolated tenants.
Each tenant gets separate virtual environments, segmented networks and role-based access, with management traffic on its own networks. The hardware is shared and the boundaries are private.
05 /
Capacity arrives as a request.
On a pooled platform, a new workload is allocated from resource pools and provisioned through a portal. The lead time is measured in approval steps rather than hardware shipments.
The platform
The platform,
layer by layer.
Every layer of the stack and what it does, from hypervisor to facility.
Layer 01
Enterprise clustering, high availability, role-based access and advanced networking on an open-source foundation with commercial support.
Layer 02
All-NVMe, software-defined storage with data replicated across the cluster, so the loss of a node is absorbed without an outage.
Layer 03
MultiPortal provides tenant provisioning, resource allocation and VM lifecycle across clusters and data centers, the layer service providers historically leaned on Cloud Director for.
Layer 04
ModernOps orchestration handles onboarding and lifecycle workflows, LogicMonitor watches platform and network health, alerts open incidents automatically, and production changes run through testing, peer review and change-board approval.
Layer 05
Hosted in ModernOps-managed infrastructure in Valley Forge, Pennsylvania and Phoenix, Arizona, reached by carrier cross-connects or secure IPsec tunnels.
Layer 06
Backup through Proxmox Backup Server or NAKIVO patterns, reserved DR compute and storage with annual failover testing, and managed Fortinet firewalls, switching and WAN under the same accountability line.
Who does what
Division of
responsibilities.
The line between what we operate and what stays yours, in writing.
| Layer | Included in IaaS | Yours | Attachable service |
|---|---|---|---|
| Physical infrastructure, facilities, power, cooling |  | ||
| Hypervisor clustering, HA, platform health, capacity | | ||
| Tenant provisioning, VM lifecycle, virtual networking | | ||
| Platform monitoring, incident workflow, change control | | ||
| Guest operating systems and patching | Customer | Server Management | |
| Applications, middleware and data | Customer | Scoped separately | |
| Backup of protected workloads | BaaS | ||
| Disaster recovery capacity and testing | DRaaS | ||
| Firewalls, switching, WAN, wireless | Managed Network and Security | ||
| Deep telemetry and dashboards | Observability |
Access boundary
ModernOps holds infrastructure-level administrative access. Routine access to your application data stays with your team.
The old model vs ModernOps
The ModernOps
difference.
The same workloads under a different operating model.
| Dimension | The old model | ModernOps IaaS |
|---|---|---|
| Provisioning | Hardware selection, purchasing, installation, cabling | Tenant provisioning and VM lifecycle through a portal and automation |
| Infrastructure | Server, storage and network silos with duplicate idle capacity | Pooled compute, storage and networking with securely isolated tenants |
| Virtualization economics | Licensing that climbs every renewal | An open Proxmox VE and Ceph foundation priced in explicit units |
| Availability | Depends on how each environment was assembled | Clustering and HA designed into the platform, absorbed node failures |
| Tenant management | Legacy cloud-management tooling or custom scripts | Purpose-built multi-tenant, multi-cluster, multi-site management |
| Operations | Individual administrators and direct console changes | ITSM workflows, monitoring, change control, reporting, service reviews |
| Protection | Backup, DR and networking as separate projects | BaaS, DRaaS, firewall and network management attached to the same platform |
The first 30 days
The first
30 days.
Three repeatable phases from kickoff to an operating baseline.
| Phase | What happens | What you hold at the end |
|---|---|---|
| Assess and design · Days 0 to 10 | Inventory source workloads, CPU, memory, storage, network dependencies, backup requirements, availability expectations and migration constraints. Map target tenant resources, segmentation, connectivity and protection. | A validated workload inventory, target-state resource model, migration sequence and responsibility matrix. |
| Provision and migrate · Days 11 to 22 | Create the tenant, allocate compute and memory, provision Ceph-backed storage, configure virtual networking and firewall connectivity, create VMs, deploy backup agents where applicable, begin migration or restore waves. | Your isolated environment online with working compute, storage, networking, backup and administrative access. |
| Validate and optimize · Days 23 to 30 | Validate VM health, connectivity, backup jobs, monitoring, alerting, failover readiness and capacity. Establish rightsizing and the service-review cadence before ownership transitions to managed operations. | A validated platform, documented operating model, known escalation paths and a baseline for optimization. |
Scope control
Thirty days stands the environment up and validates the first workloads. A full onboarding typically runs six to eight weeks from kickoff to go-live, with discovery, migration waves and DR validation on a published plan.
The other quote
Priced
both ways.
The refresh you would own, priced beside the platform you would consume.
Because we sell hardware as well as run this platform, we price the refresh you would own next to the hosted platform you would consume, in the same spreadsheet, prepared by the same engineers. The buy-and-own version of the decision lives on the Servers & Virtualization page. IaaS covers the infrastructure layer; guest OS, patching and deeper operations attach through Server Management when you want them.
Five-year cost, illustrated Buy and refresh ModernOps IaaS
An illustration rather than a quote. The refresh curve buys peak capacity up front and leaves much of it idle, while the platform curve pays for measured units as workloads run. Your own curves come out of the working session, with the source figures attached.
Proof
Current
deployments.
Representative figures and the platform they run on.
One current onboarding.
One engineering firm’s current onboarding: 106 vCPU, 548 GiB of memory and 7,674 GiB of VM storage delivered as a managed service, with backup for 20 protected workloads, a failover-ready DR reservation, and firewall and network management attached to the same engagement.
The same platform we run internally.
The platform architecture is the one ModernOps runs its own production on: Proxmox VE clusters on Dell hardware, shared all-NVMe Ceph with replicated storage and redundant cluster interconnects.
Two regions, your choice of path.
Valley Forge, Pennsylvania and Phoenix, Arizona, reached by carrier cross-connect or secure IPsec tunnel, with monthly service reviews and annual DR testing built into the operating cadence.
Field Note · No. 01 · The exit ramp lands here
The 3-Host VMware Trap
Read the field note
Tech Talk · No. 02 · Second session, date TBD
The VMware Exit Strategy
FAQ
Frequently asked
questions.
The questions buyers ask before handing over the infrastructure layer.
01 /What is ModernOps IaaS?
A managed private-cloud platform: we operate the physical infrastructure, virtualization, storage, networking, tenant provisioning, monitoring and capacity. You run your operating systems, applications and data inside isolated virtual environments.
02 /What is the platform built on?
Proxmox VE for virtualization, clustering and HA; Ceph for distributed all-NVMe storage; MultiPortal for multi-tenant provisioning and lifecycle; ModernOps automation for orchestration and workflows.
03 /How are tenants isolated on shared infrastructure?
Separate virtual environments, segmented networks and role-based access, with management traffic on distinct networks.
04 /Where do the workloads physically live?
ModernOps-managed infrastructure in Valley Forge, Pennsylvania and Phoenix, Arizona, connected by carrier cross-connects or secure IPsec tunnels.
05 /Can you migrate us off VMware?
Yes, and it is the most common reason customers contact us about IaaS. The team that runs wave-based hypervisor migrations scopes the method per environment: discovery, target tenant build, staged waves, validation and a documented rollback path, on a published onboarding plan.
06 /Is backup included?
Backup is a separately scoped attachment: Proxmox Backup Server or NAKIVO per protected workload, with jobs monitored by the same 24/7 operation. Disaster recovery attaches the same way, as reserved compute and storage with annual failover testing.
07 /Who patches the operating systems?
Your team does, by default, because the operating systems are yours. Server Management and patching can be attached as separately scoped services if you prefer to hand that over as well.
08 /How is IaaS priced and measured?
In explicit units: vCPU, memory in GiB, VM storage in GiB and backup storage in GiB, so the invoice can be forecast from the resource allocation.
09 /What happens when something breaks or needs to change?
Alerts open incidents automatically in our ITSM system, communication follows your documented procedures, and production changes require testing, peer review and change-board approval, with an accelerated path for emergencies.
10 /Can this platform run our AI or HPC workloads?
GPU, scheduling, parallel storage and orchestration are a different architecture and are scoped separately. That conversation lives on the AI & Accelerated Compute page.
Direct to an engineer
Tell us about
your environment.
Hosts, VMs, utilization and what the current platform costs to keep. We will model the hosted platform against the refresh you would otherwise buy.