Skip to content

Reference: VM / LXC inventory on arochukwu

Snapshot 2026-05-31 evening (full read-only inventory). 23 entities total: 1 LXC + 22 VMs. 1 running (VM 189), 22 stopped. Every stopped entity's conf was last touched in a single 5-hour window on 2024-03-24 (01:07 - 06:10 CET) — built in one sitting, untouched since.

For decision context and cleanup procedure, see homelab-tracker.md Phase 4 and CLAUDE.md Section 2. For the forensic walk, see SESSION-LOG.md 2026-05-31 evening "Late-evening addendum 2: Full VM/LXC inventory captured".

Inventory table

Categories: A = on fast NFS (cleanup blocked until Phase 2). B = on local-lvm (inspectable now). C = VM 189 homeNas — DO NOT TOUCH. D = template/demo (overlaps A or B).

VMID Type Name Status Storage Disk CPU RAM net0 tag net1 tag Cat. Purpose
101 VM etcd-01 stopped fast 22.2 GB 2 4 GB 10 - A k8s cluster #1 — etcd
102 VM etcd-02 stopped fast 22.2 GB 2 4 GB 10 - A k8s cluster #1 — etcd
103 VM etcd-03 stopped fast 22.2 GB 2 4 GB 10 - A k8s cluster #1 — etcd
104 VM ctrl-plane-01 stopped fast 22.2 GB 2 4 GB 10 - A k8s cluster #1 — control plane
105 VM ctrl-plane-02 stopped fast 22.2 GB 2 4 GB 10 - A k8s cluster #1 — control plane
106 VM ctrl-plane-03 stopped fast 22.2 GB 2 4 GB 10 - A k8s cluster #1 — control plane
107 VM worker-01 stopped fast 52.2 GB 2 4 GB 10 - A k8s cluster #1 — worker
108 VM worker-02 stopped fast 52.2 GB 2 4 GB 10 - A k8s cluster #1 — worker
109 VM worker-03 stopped fast 52.2 GB 2 4 GB 10 - A k8s cluster #1 — worker
170 VM kali-blue stopped fast 32 GB 2 8 GB 60 - A red-team box (VLAN-60 era)
181 VM hap-ctrl-plane-01 stopped fast 22.2 GB 2 4 GB 70 - A k8s cluster #2 (hap-*) — control
182 VM hap-ctrl-plane-02 stopped fast 22.2 GB 2 4 GB 70 - A k8s cluster #2 (hap-*) — control
183 VM hap-worker-01 stopped fast 22.2 GB 2 4 GB 70 - A k8s cluster #2 (hap-*) — worker
184 VM hap-worker-02 stopped fast 22.2 GB 2 4 GB 70 - A k8s cluster #2 (hap-*) — worker
185 VM hap-worker-04 stopped fast 22.2 GB 2 4 GB 70 - A k8s cluster #2 — worker (VMID 185 = worker-04, not 03 — non-monotonic)
186 VM hap-worker-03 stopped fast 22.2 GB 2 4 GB 70 - A k8s cluster #2 — worker
187 VM hap-ctrl-plane-03 stopped fast 22.2 GB 2 4 GB 60 - A k8s cluster #2 — control. VLAN tag DRIFT vs siblings (which are on tag 70).
189 VM homeNas running local-lvm + 7 passthrough 32 GB OS + ~3 TB passthrough 2 10 GB 70 untagged C OMV NAS — DO NOT TOUCH. OS on local-lvm; 7 virtio disks pass-through by-id from controller; net1 added 2026-05-28 for recovery access (10.0.110.11). Auto-starts.
197 VM kali-purple stopped fast 32 GB 2 8 GB 10 - A red-team box
198 VM mgmt-devops stopped fast 32.2 GB 2 4 GB 10 60 A management box (dual-NIC, one foot in VLAN-60 era)
199 VM mgmt-workstation stopped local-lvm 32 GB 4 20 GB 10 - B management workstation. Unusually large RAM allocation suggests real workstation, not throwaway. Inspect before destroy.
500 LXC demo-centos stopped fast 20 GB 2 4 GB untagged 60 A, D CentOS demo container. Investigated 2026-05-31 — NOT a DNS server. Stopped since 2024-03-24.
8000 VM ubuntu-mini-cloud stopped local-lvm 22.2 GB 2 4 GB 60 - B, D Marked template: 1 with cloud-init source disk. Potentially useful as cloud-init template asset for Path A / Path D VM provisioning. Decide before any new VM provisioning.

VLAN tag summary

Tag Used by Meaning
(untagged) VM 189 net1, LXC 500 net0 Recovery / DHCP path. Currently the laptop-reachable path on vmbr0.
10 (SERVERS) VMs 101-109, 197, 198 net0, 199 The intended production servers VLAN per pfSense design. 12 entities total.
60 VM 170, VM 187, VM 198 net1, VM 8000, LXC 500 net1 OBSOLETE TEST ERA. Gateway 192.168.60.115. Doesn't fit the 10.0.x.x / 172.16.x.x design. 5 entities. Strong signal of pre-current-design experimental era.
70 (STORAGE) VMs 181-186, VM 189 net0 Storage VLAN. Where homeNas's NFS exports were intended to live (fast, Backup-NAS). Also where the hap-* k8s cluster #2 sat. 7 entities.
100 (mgmt) (host only via vmbr0.100) Management sub-interface on the host, no VM uses tag 100.

Storage allocation summary

Storage State Allocated Notes
fast (NFS, homeNas-exported) inactive 561.4 GB across 37 disks 18 entities. Cleanup gate.
Backup-NAS (NFS, homeNas-exported) inactive 0 GB No VM/CT disks. Storage definition exists but is empty.
local (dir on PVE root) active 0.9 GB Just VM 189's OMV install ISO.
local-lvm (LVM-thin, 348 GB total) active 86.2 GB VM 189 OS (32 GB), VM 199 (32 GB), VM 8000 (22.2 GB). Plenty of headroom for new Phase 3 infrastructure VMs.

Quick decision matrix

Decision When Affects
Bulk Category A cleanup (18 entities, ~561 GB on fast) Wait for Phase 2 network rebuild All k8s VMs + hap-* cluster + red-team + mgmt-devops + LXC 500
VM 199 inspect + decide Anytime (no blocker) ~32 GB on local-lvm
VM 8000 keep-vs-rebuild decision Before new VM provisioning starts ~22 GB on local-lvm + template-asset utility
VLAN 60 fingerprint usefulness Whenever auditing config files Distinguishes obsolete-era entities at a glance

Anomalies / things worth flagging

  • VM 187 VLAN drift: hap-ctrl-plane-03 on tag 60 while its 01/02 siblings are on tag 70. Likely the config issue that left k8s cluster #2 unfinished. Don't start it without fixing the tag (it'd come up unreachable from the rest of its cluster).
  • VMID 185 vs 186 numbering swap: VM 185 is hap-worker-04, VM 186 is hap-worker-03. Just a naming-vs-VMID mismatch, harmless.
  • VM 189 onboot: flagged onboot: 1 — auto-starts on Proxmox boot. Don't change this; it's how homeNas comes back after a reboot.
  • VM 8000 is the ONLY entity with the template: 1 flag. All the other "templates" CLAUDE.md previously referenced (the hap-* cluster) are just stopped VMs, not actual Proxmox templates.
  • VM 189 disk passthrough by-id paths to seven SCSI devices (/dev/disk/by-id/scsi-3600508b1001c...) — these are the homeNas storage arrays. Persistent across host reboots because they're addressed by serial. Do not change these without explicit Kay confirmation.