Skip to content

Verify your installation

From the dashboard

  1. Sign in using the administrator created during setup, not the Linux/SSH account.
  2. Open Overview. Resolve core identity, storage or reconciliation errors first.
  3. Open System → System Status to inspect workload and reconciliation health.
  4. Open System → Hardware to inspect each node and GPU. An optional provider without matching hardware need not be installed. Engine validation is optional and must be started explicitly; it is not a blanket certification.
  5. Check the local/public URLs in Overview or Services from the client network.

Optional administrator checks

On a managed host:

sudo systemctl status k3s --no-pager
sudo k3s kubectl get nodes
sudo k3s kubectl -n flux-system get kustomizations
sudo k3s kubectl -n identity-system get appliancesetup local

On an existing cluster, use the authorized kubectl context instead of sudo k3s kubectl. After first-run setup, expect ApplianceSetup/local to be Completed; legacy installations can report CompletedLegacy. The temporary setup route is removed.

Expected result

You can sign in, read platform status and create runtime resources. Optional modules and large model downloads may still be progressing. The next check is an actual first model response, not simply a green Pod count.

If a core component remains unhealthy, follow troubleshooting and collect bounded, redacted logs. Never paste setup claims or Secret values into an issue.