cluster-snapshot-2026-06-18

2 min read

Cluster snapshot — devata — 2026-06-18

Raw state captured via kubectl / helm / talosctl. Keep this; diff future snapshots against it.

Nodes

NAME            STATUS  ROLES           AGE   VERSION   INTERNAL-IP   OS-IMAGE          KERNEL          RUNTIME
talos-k3t-9cz   Ready   control-plane   195d  v1.34.1   192.168.1.8   Talos (v1.11.5)   6.12.57-talos   containerd://2.1.5
talos-3eq-vo2   Ready   worker          193d  v1.34.1   192.168.1.3   Talos (v1.11.5)   6.12.57-talos   containerd://2.1.5
talos-igk-9ep   Ready   worker          193d  v1.34.1   192.168.1.4   Talos (v1.11.5)   6.12.57-talos   containerd://2.1.5
talos-lqv-w4u   Ready   worker          ~25m  v1.34.1   192.168.1.9   Talos (v1.11.5)   6.12.57-talos   containerd://2.1.5   <- NEW

Capacity & live usage

NODE            CPU   RAM        MEM USED   CPU USED
talos-k3t-9cz   4     3.4 GB     60%        9%
talos-3eq-vo2   2     3.0 GB     71%        8%
talos-igk-9ep   2     3.0 GB     76%        7%
talos-lqv-w4u   16    7.4 GB      9%        1%     <- NEW (512 GB NVMe)
---------------------------------------------
TOTAL           24    ~16 GB

Helm releases

cilium             cilium-1.18.6                 kube-system
kps                kube-prometheus-stack-86.2.0  monitoring
loki               loki-7.0.0                    logging
promtail           promtail-6.17.1               logging
meshery            meshery-v1.0.16               meshery      (scaled 0)
meshery-operator   meshery-operator-v1.0.16      meshery      (scaled 0)
meshery-pg         postgresql-18.2.4             meshery-db

LoadBalancer services (MetalLB pool 192.168.1.240–250)

hubble-ui        kube-system   192.168.1.243
kps-grafana      monitoring    192.168.1.242
meshery          meshery       192.168.1.240   (scaled 0)
meshery-broker   meshery       192.168.1.241   (scaled 0)

Storage (StorageClass: local-path default, single-node)

prometheus-db    5Gi   monitoring
kps-grafana      2Gi   monitoring
storage-loki-0   5Gi   logging
meshery-pg       5Gi   meshery-db

kube-system DaemonSets

cilium         4/4   (CNI — active)
cilium-envoy   4/4
kube-proxy     4/4   (removal deferred; Cilium replaces it — safe to remove now)
kube-flannel   4/4   ⚠ REGRESSION — reappeared ~8 days ago; second CNI, should not be here

Namespaces (16)

bootcamp, cilium-secrets, default, dev, example-namespace, gateway-system,
kube-node-lease, kube-public, kube-system, local-path-storage, logging,
meshery, meshery-db, metallb-system, monitoring, networking-namespace

Gateway API

CRDs installed (gateways, httproutes, grpcroutes, referencegrants, gatewayclasses) — 0 Gateway resources defined.

Notes / anomalies at snapshot time

  • Flannel regression — kube-flannel DaemonSet is back and healthy on all 4 nodes despite roadmap Phase 0 deleting it. Two CNIs running.
  • kube-proxy still present (replacement is on in Cilium, DS removal pending Talos config — now unblocked).
  • Meshery parked at 0 replicas (PVC retained).
  • Grafana still default creds admin/homelab-admin.
  • Talos access restored: talosconfig (os:admin) verified working; full bootstrap set (controlplane.yaml, worker.yaml) on disk in ~/.