I trying to use your software PRTG network monitor. Some WMI sensors didn`t work normally, it stop with an error status with error code PE015. After moment that 60% of our sensors was crashed by this problem( we had 1600~ sensors for one probe) we try to create 2 probes for 250~ sensors for one and 600~ for second probe. But its not fixed the problem. Now its crashed on a new probes. We try to send WMI query via your WMI tester, and its work. Now we could stabilize sensors only by stop PRTG services for some minutes and run it again. Scanning interval -15 minutes.
What version of PRTG are you currently running? What type of WMI based sensors do you use the most resp. what information are you interested in? WMI is a protocol which produces high workload on the probe system and depending on the total amount of WMI based sensors, it is likely that those sensors crash with the error message described.
If you want to monitor pretty standard metrics such as CPU load, Memory and Disk consumption, Traffic, Uptime etc. I recommend to use SNMP based sensors instead of WMI based ones. Please see this article for further information.
Thank you in advance,
Hello! PRTG Version PRTG Network Monitor 184.108.40.2063 x64
We used CPU load, RAM, free HDD space, pagefile, uptime, Ethernet controller, process sensors. We dont ready to change 1600+ senrors for 110+ devices to use SNMP now.
Thank you very much for your reply.
Please note that because of the WMI protocols architecture, we officially recommend to stay below 200 WMI based sensors per probe. If you want to deploy more WMI sensors, please consider using multiple Remote Probes for load balancing or increase the scanning interval. For instance increase the scanning intervals to 10 or 30 minutes for "secondary" sensors (i.e. leave ping and other "vital" sensors in a short interval, but increase the interval for other sensors (Disks usage and other alerts which don't required immediate attention). Increasing the scanning intervals will directly impact the number of requests per second which is most likely the reason to keep your Core Server busy most of the time and encountering this issue.
Please allow me to repeat my recommendation to monitor the pretty standard metrics such as:
- SNMP CPU Load sensor
- SNMP Memory sensor
- SNMP Disk Free sensor
- SNMP System Uptime sensor
- SNMP System Traffic sensor
using SNMP based sensors. I linked the SNMP pendant sensors above.
To monitor a Windows client via SNMP you need to enable it first. Please follow our blog article on how to setup SNMP on a Windows client.
Afterwards, it is also necessary to configure the incoming SNMP communication for PRTG's point of view properly. To do so, please open the Settings tab of the device that should has the SNMP sensors and configure the corresponding settings at CREDENTIALS FOR SNMP DEVICES.
It's more likely to improve the situation than increasing the scanning interval though. If you still don't want to deploy SNMP based sensors, please decrease the number of WMI based sensors per probe and consider to deploy multiple Remote Probes as described above.
Hello again! Now i have a trouble with WMI process sensor. On some divece its crashed with errors "WMI: 80041010: The specified class is not valid." What could we use instead wmi process sensor?
You can use the above mentioned SNMP CPU Load sensor instead.
Hello! I think you misunderstand me. I need not PROCESSOR sensor, but PROCESS. Analog for this one: https://www.paessler.com/manuals/prtg/wmi_process_sensor
Oh, you're absolutely correct I overlooked that one. For monitoring a process there is only the WMI based WMI Process sensor, I'm afraid.
In this case we're coming back to the above mentioned WMI based issues that you've encountered two months ago. Did you already change something in your installation and proceeded with my suggested approaches to switch WMI based sensor to SNMP based ones? If not, I'd encourage you to do so as the error message in the process sensor is directly affected by WMI related performance issues.
Hello, again. We changed all wmi sensors exept process, service and pagefile. I found analog for sirvice sensor(dowloadable exe script sensor) , but i also need process and pagefiles sensors.
I'll raise a support ticket for you and will contact you directly via email and ask for log files that allow me to gather further insights in your installation and the problem described.
Please bear with me.