Usage monitoring
CPU, memory, disk and network charts for a VM, the small agent that collects them, and how to install or remove it.
Open a VM and look at Resource usage on its Overview tab. It shows four charts: CPU, memory, disk and network, for the last hour, 24 hours, 7 days or 30 days. Hover over a chart (or focus it and use the arrow keys) to read the exact values at a point in time. A data table is available under the charts.
Our cloud providers do not report usage from inside a VM, so these numbers come from a small monitoring agent that runs inside your VM and sends them to us.
What the agent sends
Every 60 seconds, and nothing else:
| Value | Where it comes from |
|---|---|
| CPU usage in percent | /proc/stat |
| Memory used and total | /proc/meminfo (used = total minus available) |
| Root disk used and total | df for the / filesystem |
| Network bytes in and out per second | /proc/net/dev, real network cards only |
The agent does not read your files, processes, logs, users or network traffic. It opens no port and never runs commands from us: it only sends numbers and ignores whatever comes back. It is a short shell script you can read before installing it: https://veneshcloud.ir/api/agent/venesh-metrics-agent.sh.
It runs every minute under its own unprivileged system user (venesh-metrics) through a systemd timer, and uses almost no CPU or memory.
New VMs
When you create a VM, Install monitoring agent is ticked by default. The agent is then installed automatically the first time the VM boots, and the first chart points appear about two minutes later. Untick the box if you do not want it; you can still install it later.
Install on an existing VM
- Open the VM and, under Resource usage, choose Show install command (or open Monitoring agent under the charts).
- Copy the one-line command. It contains a token that belongs to this VM only.
- Connect to the VM over SSH and run the command as root (it uses
sudo).
The agent needs a Linux VM with systemd and
curl(the installer addscurlwith apt, dnf or yum when it is missing).
Remove the agent
Run this on the VM:
sudo sh /usr/local/lib/venesh-metrics/uninstall.sh
It stops the timer and removes the agent, its token, its settings and its user. Nothing is sent after that. The charts keep the data already collected until it expires.
Tokens
Each VM has its own token, and the server only stores a fingerprint of it. A token can only add numbers to its own VM's charts; it cannot read anything. If you think a token was exposed, choose Replace token: the old one stops working at once, and the agent on the VM stops reporting until you run the new install command there.
How long data is kept
| Range | Detail | Kept for |
|---|---|---|
| 1 hour, 24 hours | one point per minute | 24 hours |
| 7 days | 5-minute averages | 7 days |
| 30 days | hourly averages | 30 days |
When the charts are empty
- Monitoring agent not installed: install it with the command from the VM page.
- Waiting for the agent: the VM has a token but has not reported yet. Check that the VM is running and that
systemctl status venesh-metrics-agent.timershows the timer as active. - Agent not reporting: the agent reported before but has stopped. The VM may be stopped, or its token was replaced, or outgoing HTTPS is blocked.
journalctl -u venesh-metrics-agenton the VM shows why.
The numbers come from inside your VM, so anyone with root on the VM could change them. They are for your own information and are never used for billing.