Does anyone know why our PRTG server would experience 100% CPU utilization after the October 2018 Windows Updates? Here is our setup:

VMware - 2 CPU x 2.66GHz, 8GB RAM, Windows Server 2016, 100GB Storage

We only have 90 sensors - all Pings - except for the Monitor Server itself.

It started with a WMI error on the PRTG server with the vmxnet3 ethernet adapter - "80041010: The specified class is not valid". I deleted that sensor, but the CPU Utilization on the server is still 100%.

Looking for any kind of help.

Thanks!

Matt

Article Comments

Hello Matt,

Thank you very much for your contact.

Please update PRTG to the latest version 18.4.45.1898 to make sure that the performance issues are caused due to some kind of memory leak, even though we are not aware of any at present.

If you still encounter the performance issues after the update, please contact us by reaching out via PRTG. You can provide us with a support bundle from your installation automatically, which will give us a first insight on what's going on in your installation.

Best regards,
Sebastian


Oct, 2018 - Permalink

Hey Sebastian,

Thank you for the information. We currently have the VM off because it was locked up and unusable (after restarting it once). I will turn it on later today and check the PRTG version. We do have PRTG configured to auto-update itself (download and install automatically), so it may already be on the latest version. I'll let you know what I find.

Thanks!

Matt


Oct, 2018 - Permalink

Hi Matt,

Ok, no problem. In some cases the auto-update takes a while to find the latest version, so please make sure to manually update if the auto-update did not take place yet.

Best regards,
Sebastian


Oct, 2018 - Permalink

Hi Sebastian,

Thank you again for your help. I was back in the office today and was able to turn the server back on and get the PRTG services running again. The first thing that it did was to upgrade to the version you specified. It seems to be running stably now, but if it starts acting up again, I'll let you know.

Thanks,

Matt


Oct, 2018 - Permalink

Sebastian,

The VM was fine earlier today, but as of about 7PM tonight, the CPU was back at 100%.

After doing some more digging on the PRTG site, I found that I had some issues with the configuration in VMware. For example, I had two CPU's on individual sockets and non-reserved memory - https://helpdesk.paessler.com/en/support/solutions/articles/760000635119193-i-run-prtg-on-vmware-how-can-i-obtain-best-performance.

Hopefully fixing some of these settings will stabilize the server!

Matt


Oct, 2018 - Permalink

Sebastian,

I am still seeing 100% CPU utilization from time to time on this VM (2-3x day) and I'm seeing a "DistributedCOM" error in Windows quite frequently (several times per hour):

EventID: 10010 The server {} did not register with DCOM within the required timeout.

Any help would be appreciated.

Thanks,

Matt


Nov, 2018 - Permalink

Hi Matt,

Thank you very much for your response.

In this case it's necessary to check the log files of PRTG for further analysis. Please contact the Paessler Support directly via PRTG. This will provide us with the system log files automatically.

We'll have a look into this and will recommend the next steps.

Best regards,
Sebastian


Nov, 2018 - Permalink