Skip to content

Proxmox

Nodes

PVE1 (Primary)

Hardware

Component Details
CPU Intel Core i3-4005U @ 1.70GHz (2 cores, 4 threads)
RAM 16GB
Boot Disk Kingston SA400S37 240GB SSD
PVE 9.1.14 (kernel 7.0.2-4-pve)

Storage

Pool Type Size Purpose
rpool ZFS 222GB Main storage pool
local Directory 183GB ISO images, backups
local-zfs ZFS Pool 174GB VM/CT disks

Network

flowchart LR
    subgraph Bridge[vmbr0 - 192.168.1.199/22]
        eno1[eno1<br/>Internal LAN]
        usb[enx3c18a0d4eddd<br/>USB Adapter]
    end
    Bridge --> GW[Gateway<br/>192.168.1.254]

PVE2 (Secondary)

  • IP: 192.168.1.151
  • URL: pve2.mdhmedia.uk
  • Role: Secondary hypervisor, heavy workloads

Hardware

Component Details
CPU 2x Intel Xeon X5670 @ 2.93GHz (12 cores, 24 threads)
RAM 72GB
Boot Disk PNY CS900 120GB SSD
PVE 9.1.14 (kernel 7.0.2-4-pve)

Storage

Pool Type Size Purpose
rpool ZFS 110GB Main storage pool
local Directory 93GB ISO images, backups
local-zfs-pve2 ZFS Pool 107GB VM/CT disks

Network

flowchart LR
    subgraph vmbr0[vmbr0 - 192.168.1.151/22]
        eno1[eno1]
    end
    subgraph vmbr1[vmbr1 - 192.168.1.152/22]
        enp4s0f0[enp4s0f0]
    end
    subgraph vmbr2[vmbr2 - 192.168.1.153/22]
        enp4s0f1[enp4s0f1]
    end
    vmbr0 --> GW[Gateway<br/>192.168.1.254]

LXC Containers

Statuses as of 2026-07-07. The cluster was rationalised that day — 13 stopped test/legacy guests (10 CTs + 3 VMs) were removed (see the note below); only the running containers remain.

VMID Name Node Status IP Purpose
101 portainer pve1 running 192.168.1.241 Docker host + Portainer (all git stacks)
112 caddy pve1 running 192.168.1.245 Reverse proxy
113 homepage pve1 running 192.168.1.242 Dashboard
114 mariadb pve1 running 192.168.1.251 Database server
116 alpine-wireguard pve1 running 192.168.1.236 VPN server (WG Dashboard on :10086)
120 adguard pve1 running 192.168.1.244 DNS filtering (active)

Cluster rationalisation — 2026-07-07

Removed these stopped guests with pct/qm destroy --purge. Those with a pre-destroy vzdump on their node's local are marked ✓; the rest were throwaway/superseded and taken without a backup:

Removed Node Was Backup
CT 102 pve1 teleport (WireGuard + Authelia OAuth supersede it) ✓
CT 103 pve1 docker0X template (stale) —
CT 104 pve1 pihole (never integrated) —
CT 105 pve1 ubuntu template (stale) —
CT 106 pve1 adguard (old, superseded by CT 120) —
CT 108 pve2 plex (dup; Plex runs on Holly .210) —
CT 111 pve1 librenms ✓
CT 115 pve1 vaultwarden (test) ✓
CT 117 pve2 scanopy (test) ✓
CT 119 pve2 n8n (LAN automations move to Jarvis) ✓
VM 100 pve1 zabbix ✓
VM 107 pve1 apiscp-dev ✓
VM 109 pve2 PMS (dup Plex) —

CT 118 (grafana) was likewise destroyed 2026-07-06. Network mapping (Zabbix/LibreNMS) is deferred to Jarvis rather than replaced. The vzdumps live on each node's local and can be pruned or offsite-archived later.

Virtual Machines

VMID Name Node Status RAM Disk Purpose
121 hermes-jarvis pve1 running 4GB 32GB (ZFS, discard=on,ssd=1) Jarvis host — Hermes agent at 192.168.1.252 (see Jarvis)
110 unraid pve2 running 32GB USB boot Unraid NAS "Holly" — 16 cores, 3× PCI passthrough, OVMF + TPM 2.0

Backups

Backups configured 2026-07-06

Two vzdump jobs exist (pvesh get /cluster/backup):

Job Guests Schedule Retention
jarvis VM 121 Sun 02:00 keep-weekly 2
priority CTs 101, 112, 114, 116 daily 02:30 keep-daily 3 + keep-weekly 2

Both snapshot mode / ZSTD → local. VM 110 (Unraid) is intentionally excluded (PCI passthrough / USB boot). Offsite: pve-offsite-sync.timer mirrors /var/lib/vz/dump nightly (~04:00) to a jailed Hetzner Storage Box sub-account (u195506-sub5, home homelab) via rclone SFTP — see proxmox/backup/ in the repo.

Backup Strategy

Backup Best Practices

Regular backups are critical for disaster recovery. Proxmox Backup Server (PBS) or built-in vzdump can be used for VM/CT backups.

Backup Type Frequency Retention Time
Full Backup (Critical VMs/CTs) Daily 7 days 02:00
Full Backup (Non-critical) Weekly 4 weeks Sunday 03:00
Configuration Backup Daily 30 days 01:00

Priority Containers for Backup

Based on the running services, prioritize backups for:

VMID Name Priority Reason
112 caddy High Reverse proxy - critical for all services
113 homepage Medium Dashboard configuration
114 mariadb High Database - contains application data
116 alpine-wireguard High VPN access - contains peer configurations
110 unraid High NAS VM - backup config only (data on array)

Storage Locations

Location Type Purpose
local Directory Short-term backup storage
[NFS_SHARE] NFS Mount Off-node backup storage (recommended)
[PBS_SERVER] Proxmox Backup Server Deduplicated long-term storage

Off-site Backups

Consider configuring off-site backups to a remote location or cloud storage for disaster recovery. Options include:

  • Proxmox Backup Server with remote sync
  • Rclone to cloud storage (Backblaze B2, Wasabi, etc.)
  • NFS mount to separate physical location

Backup Configuration

To configure scheduled backups via the Proxmox UI:

  1. Navigate to Datacenter > Backup
  2. Click Add to create a new backup job
  3. Configure the following settings:
Storage: [YOUR_BACKUP_STORAGE]
Schedule: [CRON_EXPRESSION]
Selection Mode: Include selected VMs
Mode: Snapshot (recommended for running VMs)
Compression: ZSTD (best compression/speed ratio)

Restore Testing

Test Your Backups

Schedule quarterly restore tests to verify backup integrity:

  1. Select a non-production time window
  2. Restore a backup to a temporary VM/CT
  3. Verify the restored system functions correctly
  4. Document any issues and adjust backup strategy