Documentation forOrion Platform

Troubleshoot Nutanix Hardware Health monitoring

This Orion Platform topic applies only to the following products:

SAMVMAN

Consult this section if you encounter issues when monitoring Hardware Health for Nutanix clusters and vCenter/ESX hosts. You can also check the SolarWinds Success Center and the SolarWinds online IT community, THWACK. for tips,

Before proceeding, review Nutanix Hardware Health monitoring requirements and note the following details:

  • Currently, only Hardware Health monitoring for Nutanix is supported. VMAN Recommendations and Capacity Planning reports are not supported. For details, click here.
  • To manage and troubleshoot AHV hosts, use the Nutanix Acropolis tool. The Orion Platform does not currently support AHV management.

Issue: I added a Nutanix cluster and then added some ESX hosts but I don't see related data on the Virtualization Summary page.

SolarWinds recommends adding vCenters and ESX hosts with the Add Node Wizard before adding the associated Nutanix cluster. For steps, see Add a vCenter, Hyper-V host, or standalone ESX host for monitoring in VMAN.

If you added the Nutanix cluster first and then added vCenter/ESX hosts, use the Orion Service Manager to restart the following services:

  • SolarWinds Information Service (SWIS)
  • SolarWinds Cortex service
  • SolarWinds Orion Module Engine services

Issue: Hardware Health data is not updated and the last polling time didn't change.

Querying a Nutanix cluster from multiple sources can cause a Nutanix Cluster Check (NCC) lock that prevents the Nutanix API from returning data to the Orion server. This can occur if:

  • A Nutanix ncc hardware_info command (for example, ncc hardware_info show_hardware_info) was run against a manually connected CVM device.
  • Two Orion instances are monitoring the same Nutanix cluster.

Polling may also fail if a Nutanix device automatically fires an Update command to refresh sensors.

Issue: Many Nutanix values appear as UNKNOWN in Hardware Health widgets.

The Orion Platform gathers available data from the Nutanix API. If a specific data set is not available, it appears as UNKNOWN in Hardware Health widgets.

Issue: I can't find Nutanix hardware sensor events for alerts.

Nutanix hardware events only appear in the Alert Manager if the filter is set to All Events when configuring alerts. They don't appear if you select a specific Event Type such as Hardware Sensor Warning. To learn more about alerts, see Use alerts to monitor your environment with the Orion Platform.

Issue: Some nodes appear twice in Summary widgets.

Nutanix is hypervisor-agnostic. It has proprietary hypervisors, but also supports hypervisors from other vendors such as VMware. As a result, you may see "child" records repeated as parent records in some widgets, similar to those shown in the following example:

Issue: What do the POPULATED, EMPTY, and PRESENT, OK sensor statuses mean?

POPULATED indicates that a CPU is Up and while EMPTY means the CPU is not present, which translates as Unknown in the Orion Platform. PRESENT, OK indicates a PSU was detected. See Understand Nutanix sensor status mapping.