diff --git a/infrastructure/core/node-maintenance.yaml b/infrastructure/core/node-maintenance.yaml index d3eca0d0..e2a0c094 100644 --- a/infrastructure/core/node-maintenance.yaml +++ b/infrastructure/core/node-maintenance.yaml @@ -5,7 +5,13 @@ apiVersion: v1 kind: Node metadata: name: titan-14 + labels: + hardware: rpi4 + node-role.kubernetes.io/worker: "true" + longhorn-host: "true" annotations: + # Keep the runtime's other node fields; quarantine only owns placement. + kustomize.toolkit.fluxcd.io/ssa: Merge kustomize.toolkit.fluxcd.io/prune: disabled atlas.bstein.dev/maintenance-reason: "Runtime USB flash saturation; replace or repair media and verify I/O before uncordoning." spec: @@ -15,7 +21,12 @@ apiVersion: v1 kind: Node metadata: name: titan-11 + labels: + hardware: rpi5 + node-role.kubernetes.io/worker: "true" + longhorn-host: "true" annotations: + kustomize.toolkit.fluxcd.io/ssa: Merge kustomize.toolkit.fluxcd.io/prune: disabled atlas.bstein.dev/maintenance-reason: "Repeated fresh undervoltage events; verify power delivery under load before accepting new workloads." spec: @@ -25,7 +36,12 @@ apiVersion: v1 kind: Node metadata: name: titan-18 + labels: + hardware: rpi4 + node-role.kubernetes.io/worker: "true" + longhorn-host: "true" annotations: + kustomize.toolkit.fluxcd.io/ssa: Merge kustomize.toolkit.fluxcd.io/prune: disabled atlas.bstein.dev/maintenance-reason: "Sustained runtime I/O stalls prevent Longhorn engine startup; reduce load and verify storage before uncordoning." spec: