VMware Host Hardware Status (SOAP) sensor - Error 1 element returns an error state: Memory

We're receiving persistent PRTG alerts for the VMware Host Hardware Status (SOAP) sensor on our VMware ESXi hosts, with the error message Error. 1 element returns an error state: Memory. How can we resolve this?

It may be that the ESX host reports a wrong status in its SOAP-API (but does show the correct status directly in the interface). We see this every now and then sadly.

Please connect directly to the problematic ESX host with the VSphere Client and restart the 'vxpa' service.

You can also try to restart the CIM service on the ESX as well.

To add another component for complexity, please note that PRTG version 20.1.57.1745 also updated the VMware Host Hardware Status (SOAP) sensors. They now categorize unknown states reported by the VMware host server as entity with its own sensor channel. Please note that an unknown state reported by VMware isn't a problem per se. It can mean that the VMware host is not able to provide information for a certain hardware states. For instance, a physically available memory slot which is currently unequipped would be reported as unknown state.
Now, the available sensor setting for Known Warnings is only purposed for actual warning states of the host server. Because of this, you encounter the sensors to show a new warning state after the PRTG update. In this case, please switch the Ignore Unknown States setting within the sensor's Settings tab from the default option to Do not show unknown states.