Use the config_copy.yaml file to set the environment.
Change the name of the config_copy.yaml file to config.yaml (system is still in developing and has hardcoded the name of the file to load the Proxmox configurations).
Currently, system collects only the following metrics for the Proxmox nodes:
- Total number of CPUs of the PVE node (number)
- Status of the PVE node (1 if online, 0 for offline)
- Total RAM memory used on PVE node (bytes)
- Total CPU usage of the PVE node (seconds)
Exporter can retrieve metrics from all nodes within the cluster, adding the necessary label value regarding this.
When is only a single Proxmox node, the value for the label "cluster" is "", which means is required to remove it manually on the Prometheus job definition in relabel_config or (like also read in internet), labels without value are removed automatically by Prometheus.
The Exporter is developed taking in consideration that:
- A cluster contains 1 or more PVE nodes
- A PVE node contains 0 or more VMs
- A PVE node contains 1 or more storage systems
Labels include this hierarchy for easy querying. Updates will be provided.
Use the provided Dockerfile in this repository.
Run: docker build -t pve:latest .
You may use the docker-compose.yml file to create the container.
Locally
python exporter.py
Open browser at localhost:8000/metrics to populate the metrics.
Container
Open browser at docker_host_ip_address_ordns_name:8000/metrics to populate the metrics.