Warning
License: Pro - Requires a Pro or Enterprise license.
Overview¶
The Hyper-V HA probe monitors Hyper-V Replica and Windows Failover Cluster health on a Windows host. It reads two WMI namespaces locally — root\virtualization\v2 for replica relationships and root\MSCluster for cluster node and resource-group state — so no credentials or network endpoint are configured. One probe instance monitors the host it runs on.
Collected data:
- Probe reachability (the replica WMI namespace answered this cycle)
- Per-VM Hyper-V Replica health, raw replication state, and replication lag
- Failover Cluster node state (per node)
- Failover Cluster resource-group state (per group)
All metrics are emitted under the senhub.hyperv_ha.* namespace. Per-VM,
per-node and per-group series each carry a resource attribute (vm.name,
cluster.node, cluster.group) that also acts as a filter in the Web UI Sensor
Builder.
The Failover Clustering feature is optional: when it is not installed the
root\MSCluster query fails and the probe simply emits no cluster metrics for
that cycle, without erroring out. Replica metrics are unaffected.
Quick Start¶
Basic Configuration¶
# probes.d/40-hyperv-ha.yaml — each file under probes.d/ is a YAML array of probes
- name: hyperv-ha-local
type: hyperv_ha
params:
interval: 60
The probe queries WMI on the local host under the agent's service account. There is no endpoint, username or password to configure — the agent must run on the Hyper-V host itself.
Longer Interval¶
Replica and cluster state change slowly; a longer interval keeps WMI load low:
# probes.d/40-hyperv-ha.yaml
- name: hyperv-ha-local
type: hyperv_ha
params:
interval: 300
timeout: 30
Configuration Parameters¶
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
interval |
integer | No | 60 |
Collection interval in seconds |
timeout |
integer | No | 30 |
WMI query timeout in seconds |
Metrics Collected¶
Every datapoint carries a metric_type tag. Per-resource series additionally
carry the VM, node or group identifier as an OTel attribute so instances stay
distinct in OTLP/Prometheus.
Overview¶
| Metric | Unit | Description |
|---|---|---|
senhub.hyperv_ha.up |
1 |
1 when the replica WMI namespace answered this cycle, else 0 |
Replica¶
| Metric | Unit | Resource attribute | Description |
|---|---|---|---|
senhub.hyperv_ha.replica.health |
1 |
vm.name |
Replication health (1 = Normal, 0 = Warning/Critical) |
senhub.hyperv_ha.replica.state |
1 |
vm.name |
Raw replication state numeric value |
senhub.hyperv_ha.replica.lag |
s |
vm.name |
Seconds since the last successful replication |
Cluster¶
| Metric | Unit | Resource attribute | Description |
|---|---|---|---|
senhub.hyperv_ha.cluster.node.state |
1 |
cluster.node |
Cluster node state (1 = Up, 0 = Down/Paused/Joining) |
senhub.hyperv_ha.cluster.group.state |
1 |
cluster.group |
Cluster resource-group state (1 = Online, 0 = Offline/Failed/Partial) |
Filtering (Web UI Sensor Builder)¶
The PRTG/Web UI Sensor Builder exposes filters for this probe:
- Replicated VM — pick a specific replicated virtual machine
- Cluster Node — pick a specific Failover Cluster node
- Cluster Group — pick a specific Failover Cluster resource group
Requirements¶
- The agent must run on the Hyper-V host (Windows). WMI is queried locally; the probe is refused on non-Windows platforms.
- The Hyper-V role must be installed for replica metrics
(
root\virtualization\v2). - The Failover Clustering feature is optional — cluster metrics appear only
when it is installed (
root\MSCluster); otherwise they are silently skipped. - The agent service account needs WMI read access to those namespaces.
Outputs¶
Hyper-V HA metrics are available through every configured output — OTLP, Prometheus, and the pull formats (PRTG, Nagios, Web UI). For PRTG and Nagios, query the probe by its configured name: