Skip to main content
High availability is an operating model, not a checkbox. Before you choose the High Availability Kit as a target direction, confirm that the basics are in place. This guide is for the moment when “more reliable” starts to mean real customer, team, or business risk. It helps separate useful resilience planning from expensive overbuild.

​
What this guide helps you decide

  • whether downtime has real consequences
  • whether multiple nodes are enough to justify HA planning
  • whether backups, monitoring, and incident ownership are ready
  • whether Base Kit should still be the deployable first step

​
Current status

The High Availability Kit is documented as a vision-oriented target architecture. Use Base Kit for production deployments today unless you are intentionally evaluating the HA design.

​
Readiness checklist

You are closer to HA readiness when you can answer yes to most of these:
  • There are at least three suitable nodes.
  • Important services have documented recovery steps.
  • Backups are tested, not only configured.
  • Monitoring exists outside the service being monitored.
  • DNS and routing failure modes are understood.
  • Someone owns incident response.
  • You know which services actually require high availability.

​
Common false signals

These do not automatically mean you need HA:
  • owning multiple servers
  • wanting faster services
  • running public apps
  • having many Docker containers
  • wanting backups
Those needs may still fit Base Kit or Modern Homelab. The stronger signal is not infrastructure size. It is the consequence of failure.

​
When HA is justified

High availability is worth planning when downtime creates business, customer, or operational risk. Examples:
  • startup services with customer impact
  • authentication or access services required by a team
  • production apps with external users
  • critical monitoring or routing infrastructure
If no one owns response and recovery, the setup is not highly available yet. It is only distributed.

​
How this maps to the finder

The finder treats startup intent, 3+ servers, public services, and startup-service use cases as HA signals. It may still recommend “Base Kit first” as the deployable starting point while marking High Availability as the target direction.