Hyper-V¶

The hyperv probe monitors Hyper-V virtual machines on the local Windows Server host via WMI (root\virtualization\v2), reporting per-VM CPU usage, memory assignment and running state.

Windows Server only. The probe does not start on Linux or macOS.

Quick start¶

# probes.d/10-hyperv.yaml — each file under probes.d/ is a YAML array of probes
- name: hyperv
  type: hyperv

No parameter is required: the probe reads WMI on the local host. Only the collection interval can be tuned.

Parameters¶

Parameter	Must set	Default	Description
`interval`	No	`60`	Seconds between collections

Metrics¶

Metric	Unit	Description
`senhub.hyperv.up`	1	1 when the Hyper-V WMI namespace is reachable
`hyperv.vm.cpu.usage`	%	CPU utilization (0 to 100) per virtual machine, tagged with `hyperv.vm.name`
`hyperv.vm.memory.usage`	By	Memory currently used by the VM, as reported by Hyper-V
`hyperv.vm.state`	1	1 when the VM is running, 0 otherwise
`hyperv.vm.count`	{vm}	Number of VMs per state, tagged with `state` (`running`, `stopped`, `paused`)

Operational notes¶

The agent must run with administrator privileges — the Hyper-V WMI namespace is access-controlled.
One set of metrics per discovered VM; the hyperv.vm.name tag carries the VM's display name and the vmid tag its immutable identifier.
Per-VM CPU and memory are read from the summary information Hyper-V keeps for each VM; a VM without it only reports its state.

Metric reference¶

Every metric this probe can emit. Metric is the OpenTelemetry name the OTLP, Prometheus and Zabbix outputs derive theirs from. Name is what a Nagios check and the API metrics= filter match. PRTG channel is the label PRTG shows, placeholders filled from the series' tags.

Metric	Name	PRTG channel	Unit	Description
`senhub.hyperv.up`	`senhub.hyperv.up`	Hyper-V Reachability	#	1 when the Hyper-V WMI namespace is reachable, 0 otherwise
`hyperv.vm.cpu.usage`	`hyperv.vm.cpu.usage`	VM {hyperv.vm.name} CPU Usage	%	CPU utilisation of the virtual machine in percent (0–100)
`hyperv.vm.memory.usage`	`hyperv.vm.memory.usage`	VM {hyperv.vm.name} Memory Usage	Bytes	Memory consumed by the virtual machine in bytes
`hyperv.vm.state`	`hyperv.vm.state`	VM {hyperv.vm.name} State	#	1 when the VM is in running state, 0 otherwise
`hyperv.vm.count`	`hyperv.vm.count`	VM Count ({state})	#	Number of virtual machines in the given state (running / stopped / paused)