← Back to projects

Systems Engineering

Virtualized Engineering Infrastructure for Scalable Workflows

Designed a flexible engineering infrastructure around Proxmox to support PDM, service isolation, multi-OS expansion, and controlled backups under tight budget and organizational constraints.

Year

2025

Role

Engineer / System Architect

Client / Context

Internal infrastructure

Duration

Initial implementation and ongoing development

Virtualization host platformWindows and Linux service separationPDM-ready data environmentMulti-layer backup and recovery design

Context

Problem Context

The existing engineering environment relied on shared Dropbox folders, inconsistent file discipline, and no authoritative system for revision control. This worked tolerably with very small teams, but broke down as more engineers became involved: users lost track of valid revisions, older files were mistaken for current ones, and missing uploads introduced uncertainty into day-to-day engineering work.

Constraints

Constraints and Operating Conditions

  • No meaningful budget for new enterprise infrastructure beyond already-purchased hardware
  • Single-person implementation and recovery ownership
  • Need to support both Windows-based engineering systems and Linux-based infrastructure services
  • Organizational adoption lagged behind technical readiness

Decision Process

System Design and Decisions

Instead of deploying a single-purpose Windows Server only for PDM, the environment was designed as a virtualized platform around Proxmox. This allowed Windows-based PDM and SQL services to coexist with Linux-based infrastructure services while keeping workloads isolated, backups more controlled, and future ERP/PLM options open without forcing a rebuild of the underlying system.

Architecture

System Layout

System Overview

Virtualized engineering infrastructure with separated services and mixed operating environments

The platform was designed around Proxmox as a host layer to support controlled separation between engineering data systems, operational services, and future technical expansion. Rather than solving only one immediate software need, the architecture created a flexible base for Windows- and Linux-based workloads under constrained budget and uncertain long-term direction.

Host Layer

Proxmox as infrastructure control plane

Proxmox was used as the primary virtualization layer to separate roles, preserve flexibility, and reduce the risk of having to rebuild the environment later around a different ERP, PLM, or service direction.

Rack server hardware
Proxmox virtualization hostCore layer
Isolated VM / service roles
Backup and recovery checkpointsIntegrated

Reasoning

A single-purpose Windows server would have solved one narrow problem faster, but would have reduced flexibility and increased future re-architecture risk.

Windows Workloads

SolidWorks PDMVault
SQL ServerData layer
Windows ServerApplication host

Linux / Service Workloads

MonitoringZabbix
Experiments / utility servicesLinux VMs
Future system expansionOpen path

Operational Reality

The architecture provides a strong technical base, but rollout and recovery maturity remain partly constrained by organizational alignment, staffing, and operational capacity rather than by the platform structure alone.

Implementation

Implementation Sequence

01

Assess the failure modes of the existing Dropbox-based engineering data workflow.

02

Reject a single-role Windows Server design because it would solve PDM narrowly while limiting future system flexibility.

03

Use Proxmox as the host layer to support mixed operating systems and isolated service boundaries.

04

Deploy Windows Server for PDM and SQL workloads.

05

Deploy Ubuntu-based virtual machines for infrastructure services, experimentation, and monitoring.

06

Integrate NAS-based backup layers for daily, weekly, and offline recovery paths.

07

Define milestone snapshots to preserve stable recovery points throughout system setup.

Engineering Decisions

Key Design Decisions

  • Replaced ad-hoc shared-folder engineering data handling with a platform designed around an eventual system of record.
  • Used virtualization to keep Windows PDM/SQL workloads separate from Linux-based infrastructure services.
  • Created a mixed-OS architecture that preserves future ERP/PLM flexibility without forcing replatforming.
  • Implemented layered backup paths across server storage, NAS targets, offline copies, and milestone snapshots.
  • Improved operational conditions by relocating the physical server out of the working area to reduce noise and thermal disruption.
  • Accepted higher technical complexity in exchange for service isolation, controlled recovery, and long-term extensibility.

Execution

Tools and Platforms

ProxmoxWindows ServerSQL ServerUbuntu ServerDockerZabbixNAS systemsSolidWorks PDM

Outputs

System Outputs

  • Virtualized engineering infrastructure host environment
  • PDM- and SQL-ready Windows server environment
  • Linux infrastructure environment for monitoring and support services
  • Layered backup and recovery baseline

Outcome

Result and Impact

The resulting architecture established a structured foundation for engineering data management, backup layering, monitoring, and future service expansion. Even before full organizational rollout, it replaced an ad-hoc infrastructure model with a platform capable of supporting more controlled engineering operations.

Limitations and Lessons

Limitations and Lessons

  • Revision-control problems are often system-design problems, not just user-discipline problems.
  • A single-purpose server can be the wrong answer when future system direction is still uncertain.
  • Virtualization increases complexity, but can reduce architectural rework when system requirements are still evolving.
  • Recovery capability depends not only on backups, but also on operational capacity and ownership.

Next Step

Need to structure or implement a similar system?

This project reflects an engineering approach centered on system structure, operating constraints, and long-term usability. If you are working through a similar infrastructure, workflow, or remote engineering challenge, get in touch.