Frequent WMI Calls to Windows System - Combine the calls | Feature Requests

Henry Steinhauer · 2022-09-08T22:00:13Z

We are primarily a windows server datacenter and we have noticed that when monitoring the window systems, there are spikes caused by the high number of unique WMI calls to each system being monitored. Using the debug !tlist ip-address we saw that there were cases where we had more than 100 unique calls being done to gather data. This often causes other system engineers to think that LogicMonitor is the cause of the issue because suddenly there is a queue of processing on the CPU because of so many tasks coming at the system. This became very noticeable when we added the community module to monitor the running processes. That gives a count of the number of unique processes and memory used, but does not show the CPU usage of those processes. The WMI service on the system was showing excessive queueing of tasks and was spotted when another engineer was on the live system using Tsk Manager. I hate it when the monitoring tool appears to be the service that is causing the issue. Could this be cleaned up with a better process of gathering this data? Even the standard queries- if it is an 8 cpu box - there are 8 wmi calls to gather the same information except it is a different CPU number. The same is true for each of the disk letters - a unique WMI call for each disk that is defined.

Discussions

Frequent WMI Calls to Windows System - Combine the calls | Feature Requests | LogicMonitor Community