← All projects

PROJECT / 01 — PLATFORM ENGINEERING

Project Pathfinder

A documented, automated and observable self-hosted platform. Building systems I can understand — and recover when things go sideways.

  • Linux
  • Automation
  • Observability

01 / THE IDEA

What is it?

Pathfinder brings my Linux and homelab work together as a platform-engineering project. The focus is not simply running services: it is understanding how the pieces fit together, keeping the decisions documented, and making day-to-day operations repeatable.

02 / THE DIRECTION

What I’m aiming for

  • Document the architecture and the reasons behind each decision.
  • Automate repeatable setup and maintenance tasks.
  • Make service behaviour visible through useful monitoring.
  • Develop backup and recovery procedures that can be tested and understood.

03 / SMALL ENOUGH TO UNDERSTAND

A platform with a purpose

Pathfinder is where I bring Linux, automation and self-hosting together. I want a platform I can explain from end to end: what each part does, why it exists, and how I would put it back together after a failure.

The foundations bring together a Linux workstation, an always-on virtualisation platform, Docker services, dedicated automation and monitoring, and local backup infrastructure. Each has a clear responsibility. The workstation can be switched off without taking the monitoring with it.

How I build

  • Document reality. Keep what works separate from what is still planned.
  • Use what I have. Understand the available hardware before buying more.
  • Automate repetition. Make routine changes readable, repeatable and safe to run again.
  • Verify the outcome. Check the service, not just the command’s exit status.
  • Design for recovery. Understand the restore path as well as the backup job.

Working foundations, room to learn

The first monitoring milestone covers host metrics, dashboards and tested alert delivery. The automation baseline has a recorded milestone across virtual machines, containers and an ARM64 Pi. The lab notes below explain those decisions and their limits.

Recovery testing, offsite replication and broader service visibility are areas to develop and verify. Kubernetes belongs in the Borg Lab as a learning project before it becomes a production dependency. A roadmap item is an intention, not evidence that the work is complete.

Notes from the build