VM disk overview

[geroldmanders]

First of all, love your product. My Proxmox setup is much more VM-heavy than LXC-heavy. For LXCs Pulse is able to show how full the included drive is. For a VM, just the size of the configured drive.

As I have no clue about the difficulty, forgive me for asking if this LXC feature can be added to VMs as well?

[rcourtman]

hey @geroldmanders, just implemented this! VMs with QEMU guest agent installed now show actual disk usage just like LXCs do.

The implementation:

• Queries the guest agent get-fsinfo endpoint when available
• Aggregates disk usage from all real filesystems (skips special mounts)
• Falls back to showing allocated size when agent isn't available
• Added VM.Monitor permission to the auto-setup script

So now you'll see:

• With agent: "1.74GB / 9.13GB (19.1%)" - actual usage
• Without agent: "10GB disk" - allocated size only

It's live in the latest commit. Thanks for the suggestion!

[RLSinRFV]

@rcourtman can you please explain the process for implementing the VM Disk Monitoring Feature? Per your instrucitons, I removed the PVE node from Pulse, then tried to use the Quick Token Setup. The following script failed to run in the PVE console:

curl -sSL "http://192.168.1.65:7655/api/setup-script?type=pve&host=https://192.168.1.61:8006&pulse_url=http://192.168.1.65:7655&backup_perms=true" | bash

Is ther another way to enable the VM Disk Monitoring functionality?

I think I just need instructions to add the VM.Monitor permission to my API Token. I have everything else taken care of ("QEMU Guest Agent" enabled in the VM).

[rcourtman]

hey @RLSinRFV, what did the script output when it failed? Can you paste what you saw? You ran it in the shell on the proxmox host?

I just ran it now:

root@delly:~# curl -sSL "http://192.168.0.123:7655/api/setup-script?type=pve&host=https%3A%2F%2F192.168.0.5%3A8006&pulse_url=http://192.168.0.123:7655&backup_perms=true" | bash

Pulse Monitoring Setup for Proxmox VE

Checking for existing Pulse monitoring tokens from this Pulse server...

⚠️ Found 2 old Pulse monitoring token(s) from this Pulse server (192-168-0-123):

• pulse-192-168-0-123-1755860208
• pulse-192-168-0-123-1755902149

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🗑️ CLEANUP OPTION
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Would you like to remove these old tokens? Type 'y' for yes, 'n' for no:
y

Removing old tokens...
✓ Removed token: pulse-192-168-0-123-1755860208
✓ Removed token: pulse-192-168-0-123-1755902149

Creating monitoring user...
User already exists
Generating API token...
✅ Token created for Pulse monitoring

🔄 Attempting auto-registration with Pulse...

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🔐 SETUP CODE REQUIRED
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Enter your 6-character setup code from Pulse UI:
C2DPJ7

✅ Successfully registered with Pulse!

Setting up permissions...
Setting up PulseMonitor role for guest agent access...

✅ Setup complete!

[RLSinRFV]

Output: bash: line 1: {error:Authentication required}: command not found

I'm running this in the PVE host console as root.

[rcourtman]

hey @RLSinRFV, found the issue! The setup script endpoint was incorrectly requiring authentication. Just fixed it in the latest commit.

The script should now work without that "Authentication required" error. You'll still need to:

1. Get a setup code from the Pulse UI first (when adding a new PVE node)
2. Either enter it when prompted or provide it via PULSE_SETUP_CODE=XXXXXX

This fix will be in the next release. For now, if you just need the VM.Monitor permission added to your existing token, you can run:

pveum acl modify / -user pulse-monitor@pam -role PVEVMUser

thanks for reporting this, helped catch a bug in the auth logic

[RLSinRFV]

@rcourtman Quick Token Script now working nicely, but I'm not getting any VM disk usage. I have QEMU Guest Agent installed/enabled in my VM container and based on "df -h", it should be reporing about 65% available. Does something else need to activated/enabled?

I'm running PVE v9, Pulse 4.7.4.
Happy to try a few things if you have suggestions.

@rcourtman you've had a busy day ;-)
THANK YOU for all the great work on Pulse!

[RLSinRFV]

Updated to 4.8.0-rc.1
Still no VM disk usage reported.
Attached is sanitized diagnostic if helpful.

pulse-diagnostics-sanitized-2025-08-24.json

[RLSinRFV]

@rcourtman - went back and reviewed your notes from the last few releases. Although I do have qemu-guest-agent installed in my VM container, and have it turned on in the VM container Options, I did try to issue this command:

 sudo systemctl enable qemu-guest-agent

and receive the following results. Not sure if this means anything to you.

Synchronizing state of qemu-guest-agent.service with SysV service script with /lib/systemd/systemd-sysv-install.
Executing: /lib/systemd/systemd-sysv-install enable qemu-guest-agent
The unit files have no installation config (WantedBy=, RequiredBy=, Also=,
Alias= settings in the [Install] section, and DefaultInstance= for template
units). This means they are not meant to be enabled using systemctl.

 Possible reasons for having this kind of units are:
 • A unit may be statically enabled by being symlinked from another unit's
   .wants/ or .requires/ directory.
 • A unit's purpose may be to act as a helper for some other unit which has
   a requirement dependency on it.
 • A unit may be started when needed via activation (socket, path, timer,
   D-Bus, udev, scripted systemctl call, ...).
 • In case of template units, the unit is meant to be enabled with some
   instance name specified.

I have rebooted the VM container, but still have no VM disc usage stats in Pulse.

[RLSinRFV]

@rcourtman - Read your v4.8.0-rc.2 comments, thank you.

[RLSinRFV]

Error received when trying to install v4.8.0-rc.2:

root@pulse:/# curl -fsSL https://raw.githubusercontent.com/rcourtman/Pulse/main/install.sh | bash -s -- --version v4.8.0-rc.2

       Pulse Installation Script

=================================================

[INFO] Pulse v4.8.0-rc.1 is installed but not running
main: line 647: rc: unbound variable
root@pulse:/#

[rcourtman]

Thanks, should be fixed now

[RLSinRFV]

BTW, I did try root@pam permissions, both with a token with full priveledges as well as the a direct login (root@pam with root password) and still no VM disk stats. I think I'll just stay on v6.7.6 for the time being. Crazy that PVE v9 changed the access for this.

[ph-pulse]

@rcourtman First off, thank you for building Pulse—it's a fantastic tool for monitoring Proxmox clusters. We've deployed Pulse in our environment (Proxmox VE 8.3.1 with 5 compute nodes and 3 Ceph storage nodes), and it's working well overall. The auto-registration and API token setup completed successfully, and metrics are flowing in.

However, we’ve encountered an issue: disk usage is not showing for our VMs, even though the following steps have been completed:

• QEMU Guest Agent is installed and running inside each VM (Ubuntu 22.04).
• Guest Agent is enabled in the VM options via Proxmox GUI.
• VMs have been restarted after enabling the agent.
• get-fsinfo returns valid mountpoint data when run via qm guest cmd.
• Pulse was registered using the setup script:

  curl -sSL "http://192.168.200.200:7655/api/setup-script?type=pve&host=https%3A%2F%2F192.168.100.100%3A8006&pulse_url=http://192.168.200.200:7655&backup_perms=true&auth_token=GF84TR" | bash

We’d appreciate any guidance on what might be missing or misconfigured. If there's a specific permission or endpoint Pulse relies on to fetch disk usage, we’re happy to adjust accordingly.

Thanks again for your work - Pulse has been a huge help in visualizing our infra health.

🧩 Cluster Overview

• Proxmox VE Version: 8.3.1
• Compute Nodes: 5
• Storage Nodes (Ceph): 3
• Monitoring Tool: Pulse (Docker-based)
• VM OS for Pulse: Ubuntu 22.04
• Issue: Disk usage not showing for VMs despite QEMU Guest Agent being installed and enabled.

---

✅ Pulse VM Setup (Docker)

1. Create VM for Pulse on any compute node.
2. Install Docker inside the VM:

   sudo apt update
   sudo apt install -y docker.io

3. Run Pulse container:

   docker run -d -p 7655:7655 -v pulse_data:/data rcourtman/pulse:latest

---

🔐 API Token Setup in Proxmox

1. Go to Datacenter → Permissions → API Tokens.
2. Create a token for Pulse:
   • User: monitoring@pve
   • Token ID: pulse-token
   • Token Value: (copy this securely)

---

🧠 QEMU Guest Agent Setup (Ubuntu 22.04 VM)

1. Inside each VM:

   sudo apt update
   sudo apt install -y qemu-guest-agent
   sudo systemctl enable --now qemu-guest-agent

2. Enable QEMU Agent in Proxmox GUI:
   • VM → Options → QEMU Guest Agent → Enable
   • Reboot VM after enabling.

---

🧪 Verify Agent Response

On the Proxmox host:

qm guest cmd <VMID> get-fsinfo | jq

total-bytes and used-bytes, able to read disk usage.

---

🚀 Pulse Auto-Registration Script

Run this on one Proxmox host:

curl -sSL "http://192.168.200.200:7655/api/setup-script?type=pve&host=https%3A%2F%2F192.168.100.100%3A8006&pulse_url=http://192.168.200.200:7655&backup_perms=true&auth_token=GF84TR" | bash

Output Summary:

============================================
  Pulse Monitoring Setup for Proxmox VE
============================================

✅ Token created for Pulse monitoring  
✅ Successfully registered with Pulse  
✅ Setup complete!

[ph-pulse]

VM Disk Usage is now working in Pulse 4.15.0-rc.1 using both Username/Password authentication and API Token. I'm running Proxmox VE 8.3.1.

However, I noticed that Disk Read and Write metrics show 0 B/s. Is there any configuration I might have missed?

[ph-pulse]

disk read and write is now working in Version: 4.15.0-rc.2