Zero-Downtime Infrastructure Migration
for a Regional Healthcare Provider
Ageing on-premise infrastructure with compliance risk and no redundancy
A regional healthcare provider operating across three sites was running critical clinical and administrative systems on end-of-life Windows Server hardware. With a major software audit approaching and growing concerns around data protection obligations, their IT director needed a full infrastructure overhaul — without disrupting clinical operations for a single hour.
- End-of-life servers with no manufacturer support, creating compliance exposure under healthcare data regulations
- No virtualisation layer — every service ran on bare metal, meaning any hardware failure caused immediate outages
- Backup systems had not been tested for recovery in over 18 months; RPO and RTO were undefined
- Three geographically separate sites with no unified management plane or monitoring
- Internal IT team of two — not resourced to run a migration of this scale without external expertise
The client had received quotes from two larger MSPs. Both proposed extended migration windows of 4–6 months and required scheduled downtime blocks. Neither was acceptable given the clinical environment.
Phased migration with live workload replication — no maintenance windows required
GBWise proposed a parallel-environment strategy: build the target Proxmox cluster alongside the existing infrastructure, replicate live workloads, validate, then cut over service-by-service with instant rollback capability at every step.
Infrastructure audit & compliance mapping
Full inventory of existing hardware, services, and data flows across all three sites. Every workload was classified by criticality and mapped against the client's data protection obligations to define migration sequencing.
Proxmox cluster deployment
Three-node Proxmox VE cluster deployed at each site with Ceph storage replication. High availability configured so any single node failure triggers automatic VM restart within 30 seconds — eliminating the single points of failure that existed in the old environment.
Live workload replication & validation
Non-critical workloads migrated first. Each VM was replicated live to the new cluster, validated against the production environment, and signed off by the client's internal team before the cutover. Critical clinical systems were migrated last, with tested rollback procedures rehearsed in advance.
Backup architecture & DR testing
Proxmox Backup Server deployed with defined RPO of 1 hour and RTO of under 30 minutes. Full DR restore tested and documented before project sign-off. Automated backup verification configured to run weekly.
Centralised monitoring & handover
Unified monitoring across all three sites deployed, with alerting configured to the client's internal team and GBWise's NOC. Full knowledge transfer and runbooks provided. Ongoing managed services engaged for Tier 2/3 escalations.
Stack
(Name withheld at client request)
Outcomes at 90 days post-migration
The client subsequently engaged GBWise for ongoing Tier 2/3 infrastructure support, with a 15-minute response SLA for critical incidents. In the 90 days following migration, GBWise responded to three P1 incidents — average response time: 8 minutes.
Running a similar migration?
We offer a free infrastructure audit for qualified prospects. No commitment required.