Memory usage in Proxmox can be misleading if you don’t monitor the real values.
In this tutorial, you will learn how to monitor real Proxmox RAM usage using PRTG Network Monitor.
This guide is designed for system administrators who want accurate memory visibility.
We will configure sensors to track actual RAM consumption instead of cached memory.
PRTG displays real-time graphs and sends alerts when memory thresholds are exceeded.
With proper RAM monitoring, you can prevent slow performance and VM crashes.
This setup works well for both home labs and production Proxmox servers.
This video is part of the PRTG Network Monitoring Tutorial Series.

Step 1: Create scripts (SNMP installed in previous video)

SCRIPT RAM AVAILABLE (%)

nano /usr/local/bin/mem_available_pct.sh

#!/bin/bash
MEM_TOTAL=$(awk ‘/MemTotal/ {print $2}’ /proc/meminfo)
MEM_AVAILABLE=$(awk ‘/MemAvailable/ {print $2}’ /proc/meminfo)

if [ “$MEM_TOTAL” -eq 0 ]; then

echo 0
else
echo $(( MEM_AVAILABLE * 100 / MEM_TOTAL ))
fi

chmod +x /usr/local/bin/mem_available_pct.sh

Step 2: DECLARE SNMP EXTEND (EXACTLY THE SAME AS CPU TEMP)

Edit:
nano /etc/snmp/snmpd.conf
Add:
extend memAvailPct /usr/local/bin/mem_available_pct.sh

Restart:
systemctl restart snmpd

Step 3: GET THE OID (THE CORRECT WAY – DON’T GUESS)

Copy this exact OID (the one with memAvailPct):
1.3.6.1.4.1.8072.1.3.2.3.1.1.11.109.101.109.65.118.97.105.108.80.99.116

Step 4: Add sensor (SNMP Custom Advanced)

Channel: RAM Available
OID
Copy this exact OID (the one with memAvailPct):
1.3.6.1.4.1.8072.1.3.2.3.1.1.11.109.101.109.65.118.97.105.108.80.99.116

Unit: Percent

Recommended Warnings

• Lower Warning: < 25 %

• Lower Error: < 15 %