Proxmox Monitoring with Prometheus and Grafana: Full Stack Setup

A Proxmox cluster without monitoring is flying blind — you find out a disk filled up or a node overheated when something breaks, not before. This post walks through deploying a full observability stack that fixes that. The division of labor is worth understanding up front: Prometheus is the collector, quietly scraping numbers (CPU, RAM, disk, temperatures) from every node on a 15-second heartbeat and storing them as time series; Grafana is the face, turning those series into dashboards you can actually read; and Grafana’s alerting engine watches the same numbers so you get an email when something crosses a line — ideally hours before it becomes an outage.

Stack architecture

All monitoring services run in a single LXC container (CT 105) on pvelab02, keeping the stack isolated and easy to back up.

Task 1: Create the monitoring container

Task 2: Install node_exporter on all four Proxmox hosts

node_exporter is the little agent that exposes a machine’s hardware and OS metrics for Prometheus to scrape. Run it on each Proxmox host directly — not inside containers. The reason: containers see a virtualized slice of the system, so an exporter inside one can’t read the host’s real CPU temperature, disk health, or total memory pressure. Only the bare metal has the full picture.

apt-get update
apt-get install -y prometheus-node-exporter
systemctl enable --now prometheus-node-exporter

curl http://localhost:9100/metrics | grep node_cpu_seconds | head -5

curl -s http://localhost:9100/metrics | grep hwmon | grep temp

Task 3: Set up pve-exporter for Proxmox API metrics

apt-get update && apt-get install -y python3-pip python3-venv
python3 -m venv /opt/pve-exporter
/opt/pve-exporter/bin/pip install prometheus-pve-exporter

# Create config
cat > /etc/pve-exporter.yml {'<<'} 'EOF'
default:
user: prometheus@pve
token_name: prometheus
token_value: YOUR_TOKEN_SECRET_HERE
verify_ssl: false
EOF

cat > /etc/systemd/system/pve-exporter.service {'<<'} 'EOF'
[Unit]
Description=Proxmox VE Prometheus Exporter
After=network.target

[Service]
ExecStart=/opt/pve-exporter/bin/pve_exporter --config.file /etc/pve-exporter.yml
Restart=always

[Install]
WantedBy=multi-user.target
EOF

systemctl enable --now pve-exporter

Task 4: Install and configure Prometheus

apt-get install -y prometheus

2Configure scrape targets5 min

Replace /etc/prometheus/prometheus.yml:

global:
scrape_interval: 15s
evaluation_interval: 15s

scrape_configs:
- job_name: 'node'
  static_configs:
    - targets:
        - '10.0.0.70:9100'  # pvelab01
        - '10.0.0.71:9100'  # pvelab02
        - '10.0.0.72:9100'  # pvelab03
        - '10.0.0.73:9100'  # pvelab04

- job_name: 'pve'
  static_configs:
    - targets: ['10.0.0.68:9221']
  metrics_path: /pve
  params:
    module: [default]
    cluster: ['1']
    node: ['1']

systemctl restart prometheus
curl http://localhost:9090/-/healthy

Then open http://10.0.0.68:9090/targets and confirm every scrape job shows UP — this page is the first place to look whenever a graph goes blank:

http://10.0.0.68:9090/targets

Task 5: Install Grafana and import dashboards

apt-get install -y apt-transport-https software-properties-common
wget -q -O /usr/share/keyrings/grafana.key https://apt.grafana.com/gpg.key
echo "deb [signed-by=/usr/share/keyrings/grafana.key] https://apt.grafana.com stable main" \
> /etc/apt/sources.list.d/grafana.list
apt-get update && apt-get install -y grafana
systemctl enable --now grafana-server

3Import the two essential dashboards3 min

Dashboards → New → Import. Import by ID:

Dashboard	ID	What it shows
Node Exporter Full	1860	Per-host CPU, RAM, disk, network, temperatures
Proxmox via Prometheus	10347	VM/CT resource usage, cluster storage

http://10.0.0.68:3000/d/rYdddlPWk/node-exporter-full

Task 6: Configure alerting

[smtp]
enabled = true
host = smtp.gmail.com:587
user = your@gmail.com
password = YOUR_APP_PASSWORD
from_address = your@gmail.com
from_name = MyOfficeLab Grafana
startTLS_policy = MandatoryStartTLS

Monitoring is now live

You have:

• ✅ Real-time metrics from all four Proxmox hosts
• ✅ Cluster-level VM/CT stats via pve-exporter
• ✅ CPU, RAM, disk, and temperature visibility
• ✅ Email alerts for node-down, high CPU temp, low RAM, and full disks

The next post covers Tailscale: making every cluster service accessible remotely through a single subnet router, without opening any firewall ports.

Related posts:

• Node Exporter + pve-exporter: Complete Proxmox Metrics in Grafana — the metrics layer this alerting stack sits on top of
• Building a 4-Node Proxmox Cluster on HP EliteDesk Mini PCs — the cluster this monitoring stack watches
• Proxmox Security Hardening — lock down the cluster before exposing Grafana externally
• Tailscale Subnet Router: Remote Access to Your Entire Homelab — access Grafana from anywhere without opening ports
• Uptime Kuma: Dead-Simple Homelab Monitoring — the “is it up?” layer that complements these deep metrics
• Homepage Dashboard — surface Grafana and service health on one start page
• One UPS, Whole Homelab: Safe Auto-Shutdown with NUT — graph UPS battery charge, then trigger a safe cluster shutdown on power loss
• The Docker localhost Trap: 24 Hours of Red, Zero Real Downtime — a field note on monitors that lie from inside a container
• Real Screenshots, Zero Leaks: My Disposable Demo Lab — how the Grafana screenshots in these guides are captured safely
• Track AI Token Spend in Grafana: Claude, Codex, and Ollama — add the lab’s AI usage — tokens, latency, and real dollars — to this same stack
• Homelab Capacity Planning: What If a Node Dies Tonight? — put this history to work: simulate a node failure against the memory these dashboards recorded

Sources: Prometheus documentation, Grafana documentation, prometheus-pve-exporter.