VIRTUAL MACHINE ENERGY POWER MONITORING

🛠️ Installation Guide

This project enables energy metrics collection on virtual machines using Scaphandre, a lightweight agent in C for VM monitoring, and a custom Grafana dashboard. Follow the steps below to configure your environment correctly.

---

⚙️ Prerequisites

• Ensure the QEMU guest agent is running inside each virtual machine (VM).
• OpenNebula Frontends (FE) must be accessible and configured to support custom Prometheus collectors.

---

1. Install Scaphandre (on the physical host)

1. Download the latest release of the official Scaphandre package from GitHub and install on every OpenNebula Host:
   👉 https://github.com/hubblo-org/scaphandre/releases

2. Deploy the systemd service file for Scaphandre:

   • You can copy it from a host where Scaphandre is already installed.
   • If Scaphandre is deployed via Docker, ensure the systemd configuration matches your container setup.

3. Configure the target port:

   • Add the following environment variable in the host’s configuration:

     #/etc/default/scaphandre
     SCAPHANDRE_PROMETHEUS_PORT=9930

---

2. Deploy the C Agent (inside each VM)

Install the custom VM monitoring agent inside every VM:

1. Clone the repository:

   git clone https://github.com/6G-SANDBOX/power_monitoring

2. Build and run the agent inside the VM following the instructions provided in the get_proc_ticks folder.

---

3. Update OpenNebula Prometheus Collectors (on Frontends)

1. Copy the modified Prometheus collector files to:

   /usr/lib/one/opennebula_exporter

2. Restart the OpenNebula Prometheus exporter service:

   systemctl restart opennebula-prometheus

---

4. Import Grafana Dashboard

1. Open your Grafana instance.
2. Go to the sidebar and click Dashboards → Import.
3. Select or paste the contents of the JSON file with the dashboard definition.
4. Choose the appropriate Prometheus data source.
5. Click Import to load the dashboard.

Contributors