Troubleshoot Hardware Health monitoring in SAM for ESX host servers polled through vCenter

10/9/2018 9:24 PM

4/9/2025 8:15 PM

This guide was created to help users troubleshoot problems when polling Hardware Health information on ESX Host servers polled through a vCenter server (Management Server). 

• Sometimes a hardware sensor for hard disk, fan, or other hardware item is identified in the Orion Web Console as being in a critical or warning state. 
• Even when the item is replaced, the hardware health sensor can still display as critical or warning within the Orion Platform. This is a common occurrence when ESX hosts are polled through vCenter. SAM polls this information via the VMware API which can fail to update itself or can provide a false positive.

For example, the ESXi host shows power supply status critical:


 

Server Application Monitor

• The vCenter can fail to update or there is a significant time delay before the Hardware sensors are updated and warning messages are cleared from the VMware API.​​​​​​
• While there may are multiple causes, this issue is most likely due to polling.

To verify if the status shown for Hardware Health in the SolarWinds Platform Web Console is consistent with the status reported by the VMware API, follow this guide.

1. Verify if your ESX host is polled through vCenter or Polled Directly by checking on Virtualization settings page.

To confirm this information, go to  Settings > Virtualization Settings.

2. Access the VMware API:

• Access the VMware API by going to: https://your_vCenter_server/mob   
  • Replace your_vCenter_server with either the IP address or host name of the server to access.
• This will open the Managed Object Browser page. 

3. Locate the ESX host object.

• Select the following properties:
  • Content > rootFolder > childEntity > select datacenter > hostFolder > childEntity > select domain > host > select ESX host
• You should see page similar to this:

4. Locate the ESX host's hardware health information

• Go to runtime -> healthSystemRuntime -> systemHealthInfo -> numericSensorInfo
• You should see page similar to this:

5. Alternative paths for numeric sensors

• Except the Numeric sensors there is number of other sensor in other locations which we monitor.
• Here are alternative paths which you can substitute with the one for numeric sensors.

runtime -> healthSystemRuntime -> hardwareStatusInfo -> cpuStatusInfo

runtime -> healthSystemRuntime -> hardwareStatusInfo -> memoryStatusInfo

runtime -> healthSystemRuntime -> hardwareStatusInfo -> storageStatusInfo

hardware -> systemInfo -> otherIdentifyInfo

config -> storageDevice -> scsiLun

 

Example 1

• ESX Guest showing  "Memory Critical" issue. 
  • Please Note :  There are  more than numericSensorInfo properties.  Please check the one which is relevant to the Warning / Critical Message you are seeing to your ESX,  
• You can view the Memory Critical issue using the following properties:
  • runtime -> healthSystemRuntime -> hardwareStatusInfo -> memoryStatusInfo

The following error is displayed in SAM web console

 

However, when we check the host itself, there appears to be no issue. 

Researching further into the VMware API, we can see that it is reporting a 'Red' Status, resulting in the Warning message from the SAM Web Console

 

Example 2

• A power supply has been replaced, however the power supply status is still showing RED.

When checking the host via vCenter there appears to be no issue.

However, SAM is reporting that there is an issue with one of the Power Supply Units.

Checking the VMware API where SAM is polling, we can see there is a 'Red' warning message flagged in the API. 

Example 3

Options for Resolution

Option 1

Contact VMware directly to see if there is a resolution to 'Clear' the incorrect status's from the API. They will be able to assist you from their side as unfortunately we are unable to provide a workaround at this time for the incorrect states being provided within the API.

Option 2

Poll the ESX / ESXi directly using the CIM Protocol.