Quasarlab Docs¶
Architecture, runbooks, incidents, and command references for my homelab.
Why these docs exist¶
Two reasons. The first is selfish: when something breaks at 3am, I want to read what past me already figured out instead of rediscovering it. The second is to demonstrate, in public, how I think about systems: how I lay them out, where the failure modes are, what the tradeoffs were, and how I respond when things go sideways.
How to read it¶
If you only have a few minutes:
- Skim Architecture / Overview for the lay of the land.
- Read one incident report. They each tell a story from symptom to root cause to fix.
If you have longer:
- The runbooks are the response side of those incidents, distilled.
- The decisions explain why the lab is the shape it is.
- The cheatsheet is grep-bait for me at 3am.
What runs in the lab¶
- Three-node Kubernetes cluster on Proxmox, Calico CNI, MetalLB L2 for LoadBalancer IPs.
- ArgoCD as the GitOps engine for everything in cluster.
- kube-prometheus-stack for metrics, Loki + Vector for logs, Alertmanager into Discord via a small custom proxy.
- External Postgres VM for stateful workloads that I want to keep portable.
- Ansible for VM-level configuration, Terraform for provider-managed resources.
- Mostly homegrown apps plus the usual *arr stack, Jellyfin, ArgoCD, and a HITL Discord bridge for agentic workflows.
The full picture lives under Architecture.