HCI Server Infrastructure
The cluster under the cooperative's production billing, with a recovery plan written before it was needed.
The hyperconverged cluster that runs production billing and operational VMs, with written recovery procedures.
Team of one· Solo infrastructure architect
No team, no agency, no hand-offs. I owned this project end to end:
- Requirements
- Architecture
- Infra & deployment
- Stack
- HCIArcherOSVM OrchestrationLinuxIncident Response
What was broken, and what fixed it
Problem: Production billing and operations need infrastructure that survives hardware failure.
Fix: A hyperconverged cluster runs the production billing and operational VMs with RF=2, so data is kept in two copies and a single disk or node failure does not take it offline.
Problem: A disk failure on the billing stack needs a known recovery path.
Fix: Documented incident response procedures cover disk-failure recovery for the production billing and meter reading stack.
Approach
- Hyperconverged infrastructure with RF=2 clusters for redundancy.
- Incident response procedures written before they were needed.
What I built
- HCI cluster for production billing and operational VMs
- VM orchestration on Linux-based infrastructure
- Disk-failure recovery procedures for the production billing and meter reading stack
Architecture
Infrastructure & delivery
- HCI
- ArcherOS
- VM Orchestration
- Linux
- Incident Response
- 1LinuxHCI cluster
A hyperconverged cluster runs production billing and operational VMs on Linux-based infrastructure.
- 2
Clusters use RF=2, so data is kept in two copies across nodes.
- 3HCI cluster
VM orchestration places and manages the production workloads on the cluster.
Engineering
- RF=2 redundancy means a single disk or node failure does not take data offline.
- Incident response procedures were written before they were needed.
- A documented disk-failure recovery path covers the production billing and meter reading stack.
Outcome
- Runs the cooperative's production billing and operational workloads.