LALeonardo AboinDatabase lab notes
All applications

03 · Private network · Personal lab · 2025

Reliability lab

The platform under the other two applications.

Virtualization, two small Kubernetes distributions, backups, failure drills and a monitoring stack. There is no public URL, and there should not be one.

No public URL. The lab network stays private.

How the reliability lab is stacked
Proxmox under pfSense, Kubernetes, databases, and Prometheus Grafana LokiPrometheus · Grafana · LokiPostgreSQLMySQL · SQL ServerGeode AG2 → AGK3sMicroK8sProxmoxpfSense

What you can do in it

Pick a part of the application. The note says what it is for, what you should see, and one thing worth trying.

What it does

Proxmox

Hosts the virtual machines the rest of the lab runs on.

On the screen
A hypervisor I operate myself, rather than a rented database with the machine hidden.
Try this
There is nothing to click here. The public evidence is the two applications that sit on this platform.

What this lab is

A database is only as recoverable as the last restore someone actually performed. This lab is where I rehearse that, on hardware and virtual machines I operate myself, without waiting for the next production incident.

Apache Geode, which Multi-DB Studio uses as the in-memory grid, is in the middle of a host move from AG2 to AG. AG was repaired recently. Until that move is finished, Geode is a migration, not a settled platform.

What I practice

Backup schedules, restore drills, failure simulation, and reading a dashboard during the drill. Bash and Ansible are for the procedures I want to repeat. The runbook matters more than a clever one-off command.

What this does not prove

There is no public URL for the lab network. Grafana screenshots and a sanitized repository would make this note easier to trust. Until those exist, the evidence is the two applications above and the production work on my Upwork profile.