// Case study

GSDAC server consolidation: 16 → 2 + NAS + 12 VMs

Consolidated a 16-server production fleet into 2 hosts plus NAS storage running 12 Hyper-V virtual machines. Replication, redundancy, and failover were designed in from day one.

Infrastructure Systems Administrator — design, migration, and cutover Employer: Greater San Diego Air Conditioning (GSDAC) · 2014–2018

Tools / technologies

  • Hyper-V
  • Failover Clustering
  • NAS / replication
  • Active Directory
  • Backup & DR tooling

Problem

As the in-house Systems Administrator at Greater San Diego Air Conditioning (GSDAC), I inherited an environment that had grown to sixteen aging physical servers carrying production workloads across roles that no longer needed their own metal. The footprint was expensive to power, slow to recover, and brittle: every hardware failure became an outage, and routine maintenance windows ate into the business day.

Constraints

  • No tolerance for extended downtime during cutover — this was the business’s production environment.
  • Replication, redundancy, and failover had to be designed in, not bolted on after.
  • Hardware refresh budget was finite — the design had to last.
  • Existing services and dependencies had to keep working with no end-user retraining.

Role

End-to-end ownership as the sole systems administrator. Sized the new hosts and storage, designed the cluster and replication topology, planned the migration sequence, executed the cutovers, and ran the post-migration stabilization.

What I did

  • Sized two physical hosts plus a NAS to consolidate all 16 workloads.
  • Built a Hyper-V cluster with shared NAS storage and 12 virtual machines mapped to the legacy roles.
  • Designed replication and failover so a single host loss is survivable without business impact.
  • Phased migrations by service domain to bound blast radius at each step.
  • Validated backups and disaster recovery procedures against the new topology — not the old assumptions.

Outcome

  • 16 boxes → 2 hosts + NAS running 12 VMs.
  • Lower power and licensing footprint, fewer moving parts to maintain.
  • Faster recovery during failures; smaller and more predictable maintenance windows.
  • A baseline that absorbed the rest of the GSDAC tenure’s infrastructure work — VoIP, Exchange-to-cloud, firewall and DMZ rebuild, SCADA for the solar array — without another hardware sprint.

What it demonstrates

Comfort owning hypervisor, storage, replication, and DR as a single design surface — and willingness to make the boring, high-leverage decisions that take an environment from “fragile” to “quiet.”